Effect of chloride ions on corrosion behaviour of austenitic nickel and nickel free stainless steels in phosphoric acid solutions

Abstract

The effect of chloride ions’ presence (0·005-1·0M NaCl) in phosphoric acid solutions (5, 40 and 75%) on the corrosion behaviour of three austenitic stainless steels (an experimental steel Fe–18Cr–12Mn–0·6N and two trade grades, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N) has been studied by potentiodynamic polarisation measurements. The surface examinations of the samples tested involved X-ray photoelectron spectroscopy as well as optical and scanning electron microscopy. It was established that chlorides added to phosphoric acid solutions deteriorate the general corrosion resistance, and under anodic polarisation, they provoke pitting corrosion. The composition of the stainless steels significantly influences its corrosion behaviour in the phosphoric acid solutions containing chloride ions. The replacement of nickel with manganese and nitrogen on top of lower chromium content has a strong negative effect on the corrosion resistance.

Keywords

Stainless steel Nitrogen Chloride Pitting Phosphoric acid Polarisation

Introduction

Austenitic nickel free or of reduced nickel content stainless steels are under intensive investigations because of their low cost production and enhanced mechanical characteristics with respect to the conventional chromium–nickel counterparts. The reduction of nickel or its complete elimination in the austenitic stainless steels can be attained through its replacement by other austenite forming elements as nitrogen or/and manganese. Such steels are known as chromium–manganese austenitic stainless steels (grades 200 series), and they are encountered in many engineering applications owing to the great mechanical properties combined with acceptable corrosion resistance.^1,2

Stainless steels demonstrate good corrosion resistance in neutral and alkaline environments because of spontaneous formation of passive protective layers. Nevertheless, such steels are vulnerable to a local corrosion when aggressive ions, for instance chlorides, destroying the passive layers exist in the environment. The latter is strongly expressed when stainless steels are placed in acid environments containing halogen since most of them are highly susceptible to pitting corrosion in these conditions. Phosphoric acid solutions are one of the cases of massively produced and widely used environments which could be contaminated with chlorides.

However, systematic studies on the corrosion resistance of iron and ferroalloys and in particular of chromium–manganese–nitrogen stainless steels in acid phosphate environments containing aggressive ions are rare in the scientific literature.^3–9 Some of these works determine the effect of temperature and of various contaminants of the phosphoric acid such as Cl⁻, S^{2 −} as well as other species in small quantities emerging in the production process.^10–14 The acid concentration in these studies was around 40-50%.^11,13

In contrast, there are many sources of information about the inhibition effect of phosphate ions on the pitting corrosion of steels in chloride environments. In most of these studies, the model environments are either neutral or alkaline chloride solutions with low phosphate ion contents.^15–20 Browsing published results relevant to the effect of phosphates on the passive states of steels and consequently their pitting corrosion resistance, many contradictory outcomes can be encountered. According to some authors,^15,16,21 the phosphate ions suppress the formation of pits to a greater extent than in cases of chromate, sulphate and nitrate ions. The explanations of such strong inhibition effect are not based on a unified concept. Refaey et al.,^16,21 for example, suggest that the phosphate ions’ action is as a result of deposition of insoluble compounds on the steel surface, while Zuo et al. ¹⁵ explain the effect of the phosphate ions by a decrease in solution acidity inside the pits as a result of hydrolysis processes related to . The reduction of the local solution acidification decreases the rate of metastable pit nucleation and delays the pit growth owing to the repassivation processes. The same authors drew their conclusions from experiments in environments modelling inner pit solutions (4·13M NaCl and pH 1). The experiments revealed that, even for phosphate/chloride ratio of 1:250 in the environment, the tested stainless steel 316 remained in spontaneous passive state with increased width of the passive region. In addition, the passive current was lower than in cases when phosphates are missing. In contrast to these establishments, in similar studies on stainless steel 304 behaviour in environments containing phosphate ions of low concentration (0·014M Na₂HPO₄), it was reported by Lakatos-Varsányi et al. ¹⁹ that there were no effects of the phosphates on the current in either the active and the passive regions, and the steel pitting formation potential remained unaffected. The only reported positive effect by these authors was related to the repassivation potential, and the explanations were again based on the limitation of the pit solution acidity.

The present work takes into consideration the effect of the chloride ions’ presence in phosphoric acid solutions on the corrosion behaviour of three austenitic stainless steels: an experimental steel Fe–18Cr–12Mn–0·6N and two trade grades, i.e. AISI 304 and the Russian X14AΓ15. Precisely, the study focuses on the effect of experimental steel composition and its relationships to the exhibited corrosion resistance, compared to those of conventional nickel steel and manganese–nitrogen steel samples.

Experimental

Austenitic stainless steels Fe–18Cr–12Mn–0·6N, Fe–14Cr–15Mn–0·2N and Fe–18Cr–9Ni were used in the tests. Their chemical compositions are summarised in Table 1. All samples were thermomechanically treated under almost similar conditions to achieve austenitic structures. Oxides and oxysulphides were detected as principal non-metallic inclusions in the steel samples. The chromium–manganese–nitrogen steels are richer of sulphide inclusions than the chromium–nickel steel.

Table 1

Chemical compositions of stainless steels/wt-%

Steel	C	Si	Mn	P	S	Cr	Ni	N
Fe–18Cr–12Mn–0·6N	0·04	0·30	12·00	0·011	0·023	16·50	…	0·61
Fe–14Cr–15Mn–0·2N	0·12	…	15·00	0·015	0·020	13·45	…	0·20
Fe–18Cr–9Ni	0·055	0·52	1·28	0·022	0·009	17·49	9·37	…

Samples with a contact surface area of ∼0·5 cm² were first wet ground with 120-600 grit silicon carbon papers. Before polarisation tests, the working electrode was ground with 800 abrasive paper, then degreased in alcohol–ether mixture and washed in distilled water. Before the potentiodynamic polarisation tests, the surfaces of the samples were polarised for 5 min at − 0·7 V(SCE) in order to remove any air formed oxides.

The test environments were phosphoric acid solutions of concentrations 5, 40 and 75%, with added amounts of 0·005-1·0M NaCl prepared by dilution with monodistilled water; the chemicals used were 85 wt-% phosphoric acid and sodium chloride of class p.a.

Open circuit potential (OCP) was measured during 1 h in the test solutions, and the values at the end of the period were averaged.

Cyclic potentiodynamic polarisation measurements were performed with a potential scan rate 1 mV s^{− 1}, starting from the potential − 0·7 V(SCE) followed by anodic polarisation. After passing through the passive state and the current density exceeded one order the passive current density, a polarisation in the reverse (negative potential direction) was carried out down to the point where the curve intersected the anodic one. Every potentiodynamic polarisation test was performed five times. The results were arranged from lowest to highest value, and the middle one was considered as final. After accomplishment of the individual potentiodynamic tests, the sample surfaces were observed with an optical microscope at magnification × 200. In cases were crevices exist at the sample/isolation interfaces, the tests results were ignored.

Electrochemical polarisation curves were obtained with a galvanostat/potentiostat (Princeton Applied Research Model 263) and computer software Power Suite. The electrochemical measurements were carried out in a conventional three-electrode cell with a platinum counter electrode and a saturated calomel electrode (SCE) as a reference electrode. All potentials were reported with respect to the SCE. The experiments were carried out at room temperature in naturally aerated solution without stirring.

The film composition of Fe–18Cr–12Mn–0·6N after an anodic polarisation at 0·4 V(SCE) for 2 h in 40 wt-%H₃PO₄ (with and without addition of 0·1M NaCl) was investigated by X-ray photoelectron spectroscopy. The measurements were performed in a VG ESCALAB II system using Mg K _α radiation with energy of 1253·6 eV. The pressure in the chamber was 10^{− 8} Pa. The binding energies were determined with an accuracy of ± 0·1 eV utilising the C 1s line at 285·0 eV (from an adventitious carbon) as a reference. The changes in composition and chemical surrounding in the depth of the films were determined on the basis of the areas and binding energies of C 1s, O 1s, N 1s, Fe 2p_3/2, Cr 2p_3/2, Mn 2p_3/2, P 2p and Cr 2p_3/2 photoelectron peaks (after subtracting the background using the Shirley) and Scofield's photoionisation cross-sections.

Results and discussion

Potentiodynamic polarisation study

The effect of chloride ions on the corrosion behaviour of the steels in phosphoric acid was examined in three different acid concentrations: low, 5%; medium, 40%; and high, 75%. The typical polarisation curves are shown in Figs. 1 and 2 for Fe–18Cr–12Mn–0·6N and Fe–18Cr–9Ni steels respectively. For the tests with 75%H₃PO₄, the highest chloride concentration was 0·5M in order to avoid saturation of the solution. The polarisation relationships reveal the main electrochemical corrosion parameters providing information about: (i)

active region and general corrosion behaviour (corrosion potential E _corr and corrosion rate i _corr)

(ii)

active–passive transition (critical potential E _cr and critical current density i _cr)

(iii)

passivity and passivity breakdown (passive current density i _pass, pitting potential E _pit and repassivation potential E _rp).

Polarisation curves of Fe–18Cr–12Mn–0·6N steel in solutions of 5, 40 and 75%H₃PO₄ containing NaCl; 1 mV s^{− 1}, 20°C

Polarisation curves of Fe–18Cr–9Ni steel in solutions of 5, 40 and 75%H₃PO₄ containing NaCl; 1 mV s^{− 1}, 20°C

General corrosion behaviour

The investigation of stainless steels behaviour began with obtaining the OCP versus time dependencies. It was established that in all phosphoric acid solutions with the presence of sodium chloride, OCP moves in negative direction and stabilise in ∼5 min. This behaviour could be explained by the initially dissolving of air formed passive film followed by steel matrix dissolution. After 1 h, the samples’ surfaces were covered with dark, uneven porous layer from corrosion products. The latter indicates that steels remain in active state in this environment. There was only one exception for the samples of Fe–18Cr–9Ni steel in 5%H₃PO₄, containing chloride up to 0·1M, for which there was ennoblement of potential and clear surface after 1 h of immersion. The stationary values of OCP after 1 h stay in phosphoric acid solutions with 0·1M NaCl are mentioned on the polarisation dependencies (Figs. 1 and 2). It is obvious that OCP values are ∼50 mV more positive than the potentiodynamically measured corrosion potentials and fall within the area of active dissolution.

As a strong attack, and therefore irreproducible changes, in the steel surfaces at open circuit conditions takes place, polarisation dependencies were taken without prior stay in corrosion environment. Only short cathodic treatment was carried out, during which residues of air formed passive layers were removed without dissolution of metal surface. From as received polarisation dependencies, it was established that, in all phosphoric acid solutions containing chloride ions, all stainless steel samples tested are in active state. This can be judged from the negative values of the corrosion potential and clear active–passive transition.

The corrosion potential evolution affected by chloride ions added to the phosphoric acid solutions is presented in Fig. 3. The corrosion potentials of all samples shift in the negative direction, in accordance with the equation . The b slopes of all curves are about 20-30 mV/decade. As an exception, it can be seen that the corrosion potentials E _corr determined in solutions of 5%H₃PO₄ with chloride content up to 0·1M have relatively constant values. Similar shifts of E _corr in the negative direction with increase in chloride concentration were reported for steel 316 in 30%H₃PO₄.⁶ Moreover, in the case of OCP and 30%H₃PO₄ with 1·5%NaCl, the test with steel 316 revealed predominant dissolutions of Cr and Fe components with simultaneous enrichment of the steel surface with nickel.⁷ Under these conditions, the passive layer thicknesses were smaller than those formed in the absence of chlorides in the acid solutions.

Corrosion potential as function of NaCl concentration in 5, 40 and 75%H₃PO₄ for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels

The corrosion rates (expressed by the corrosion current density) as functions of the NaCl concentration in 5, 40 and 75% phosphoric acid solutions are shown in Fig. 4. The increase in the chloride ion concentration accelerates the general corrosion rate in acid solutions, and this effect is manifested markedly when the NaCl concentration exceeds 0·1M in 5% acid. Precisely, in a solution of 5%H₃PO₄ with added 1·0M NaCl, the corrosion current for samples of Fe–18Cr–12Mn–0·6N steel reaches its maximum of ∼8·4 Ma cm^{− 2}. This maximal corrosion current is about one order higher than that observed at low chloride concentrations and about two orders higher than that recorded in 5%H₃PO₄ in the absence of chlorides.²² Further, in 40 and 75% phosphoric acid solutions, the addition of chlorides accelerates the corrosion rate, and the corrosion current densities increase about two to three times within the studied concentration range.

Corrosion current density as function of NaCl concentration in 5, 40 and 75%H₃PO₄ for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels

Similar negative effect of the chloride ions on the i _corr is registered also by Refaey²¹ for steel corrosion in hydrochloric acid solutions. Refaey explains the results by a synergetic effect between Cl⁻ ions chemisorbed on the metal surface and H⁺ ions from the solution, which are electrostatically attracted by Cl⁻. The result is a layer of ions adsorbed to the surface acting as a catalyst for the steel dissolution reaction.²¹

Active–passive transition

The addition of chloride ions into phosphoric acid solutions hampers the transition to anodic passivity state for all steel samples tested. In general, the critical current density (Fig. 5) as well as the distinction [E _cr − E _corr] (Fig. 6) grows up. The values of i _cr and E _cr obtained with 5%H₃PO₄ scatter (see Figs. 5a and 6a), while with increase in the phosphoric acid concentration, the points depend almost linearly on the concentration (Figs. 5b and 6c) or slightly deviate from linear relationships (Figs. 5c and 6b). Similar rapid increase in i _cr and a shift of E _cr in the positive direction with increase in NaCl concentration have been reported for steel 316 in 30%H₃PO₄.⁶

Critical current density as function of NaCl concentration in 5, 40 and 75%H₃PO₄ for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels

Difference between critical and corrosion potentials as function of NaCl concentration for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels in 5, 40 and 75%H₃PO₄

It is interesting to note the fact that the presence of chlorides in the phosphoric acid affects most unfavourably the anode behaviour of the chromium–nickel steel. In solutions not containing chloride ions when anode polarisation is applied, the chromium–nickel steel is in state of spontaneous passivity (Fig. 2). At the same time, the chromium–manganese–nitrogen steels pass through active–passive transition (Fig. 1). In the presence of chloride ions in the solutions, the samples of Fe–18Cr–9Ni steel also begin to pass through active–passive transition when anodic polarisation is applied. Moreover, when the acid concentration is increased, then a well expressed peak of the anodic current for this steel is observed at lower chloride concentrations.

The critical current densities for Fe–18Cr–12Mn–0·6N steel samples are a little bit lower that those of Fe–14Cr–15Mn–0·2N steel but remain still higher with respect to the values established for the chromium–nickel steel. The difference in the values of i _cr for chromium–manganese–nitrogen steels and chromium–nickel steel is well manifested in solutions with low chloride concentrations. In 75% acid with chloride content beyond 0·1M, the samples of the three steels exhibit almost equal i _cr values.

Curiously, in pure phosphoric acid for all studied concentrations on the polarisation dependencies of chromium–manganese–nitrogen steels only, a second corrosion peak, at around − 0·4 V(SCE), was registered at the end of the active–passive area. It can be considered as a result of cathodic reaction of hydrogen evolution onto the passive surface.²³ The registered corrosion current is low and commensurable with passive current. Because of the high hydrogen ion concentration and the high environment viscosity, the participation of oxygen in this naturally aerated system is negligible. As opposed to the other authors,²³ the results received by us reveal that the chloride addition even in minimal quantity leads to formation of stable passive system and complete predominance of the anodic reaction on the cathodic one.

Passivity, passivity breakdown and repassivation

The polarisation dependencies on Figs. 1 and 2 demonstrate typical for chromium austenitic steels’ anodic passivity. When the chloride content in phosphoric acid solutions is low (up to 0·1M), the passive current density attaints minimum value at potential 0·4 V(SCE). In addition, anodic polarisation oxidation of Fe²⁺ to Fe³⁺ occurs, which results in current density increase. The state of anodic passivity of the studied steels starts at about − 0·2 V(SCE) and stretches to ∼1·0 V(SCE), where the transpassive dissolution and oxygen liberation begin. This stable passive state was observed for all three stainless steels tested (irrespective of the acid concentrations and chlorides content) below the critical value for pitting development (Figs. 1 and 2).

In all 5%H₃PO₄ solutions as well as in 40%H₃PO₄ with chloride content up to 0·1M, the passive current for Fe–18Cr–12Mn–0·6N and Fe–18Cr–9Ni steels holds in relatively stationary value of ∼10 μA cm^{− 2} (Fig. 7). The Fe–14Cr–15Mn–0·2N steel has a little bit higher passive current. In 40%H₃PO₄ with higher concentrations of active ions, the passive film protective properties considerably deteriorate. This behaviour is manifested by a sharp increase (more than two orders of magnitude) in the i _pass value. Similar result has been reported by El Dahan⁶ regarding 316 steel in 30%H₃PO₄: a sharp increase in the passive current at a concentration of 15 000 ppm NaCl (or 1·5% = 0·26M NaCl).

Passive current density as function of NaCl concentration in 5, 40 and 75% H₃PO₄ for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels

The most strongly demonstrated negative effect of the chlorides was observed in the solutions of 75%H₃PO₄, where the passive current density i _pass grew noticeably when 0·05M of NaCl was added to the solution. At higher acid and chloride concentrations, the passive current density reached ∼4 and 1 mA cm^{− 2} for chromium–manganese–nitrogen steels and the chromium–nickel one respectively without observation of pits formation. These high values of current density in the potentials corresponding to limiting current plateau indicate mass transfer controlled mechanism, in which electrochemical polishing occurs. The latter relates to high viscosity of 75% phosphoric acid combined with extreme aggression of the environment in respect to the passive layers on the steel surfaces.

In cases when the critical chloride concentration exceeds the threshold for pits formation, the passive region considerably shrinks as a result of the passive film breakdown at lower degree of anodic polarisation. The dependencies E _pit–C _NaCl for the three examined steels are shown in Fig. 8 and could be described in logarithmic coordinates, namely The constants a and b depend on the steel composition.

Passive layer breakdown potential (thick symbols and lines) and repassivation potential (dotted line) as function of NaCl concentration in 5, 40 and 75%H₃PO₄ for Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels

In 5%H₃PO₄ solutions, the pitting corrosion on Fe–18Cr–12Mn–0·6N, Fe–18Cr–9Ni and Fe–14Cr–15Mn–0·2N steels was developed when the chloride content was equal or exceeded 0·04, 0·05 and 0·01M respectively. The increase in acid concentration to 40% leads to a growth in critical chloride concentration. The stable pits development on the chromium–manganese–nitrogen steels in 40%H₃PO₄ was registered for concentration of 0·1M NaCl, but this was not enough to develop local corrosion damages when the chromium–nickel samples were tested in the same solutions. These results reveal that, with increase in the acid concentration, both the pit nucleation and growth were suppressed. The phosphate ions, however, despite the high aggressive environment with respect to the passive layers, inhibit the local corrosion activity.

The ability of the passive layer to recover itself from local breakdowns when the anodic polarisation reduces is related to the repassivation potential E _rp, determined from the cross-point of the forward and the backward branches of the polarisation curve (Figs. 1 and 2). In general, the more positive the value of E _rp, the easiest the return of the steel to the passive state. For the samples tested in this study, the repassivation ability of the passive layers in 5 and 40%H₃PO₄ deteriorates with the increase in the chloride concentration (Fig. 8); the effect is stronger in solutions of 5%H₃PO₄.

The results of potentiodynamic studies reveal that the steel's chemical composition significantly affects the corrosion resistance. In this context, the chromium–manganese–nitrogen samples demonstrate lower corrosion resistance than that exhibited by Fe–18Cr–9Ni steel samples in all tests carried out. The replacement of nickel by manganese leads to 150-200 mV more negative values of E _corr and additionally to higher corrosion, critical and passive current densities. The lower chromium and nitrogen contents in Fe–14Cr–15Mn–0·2N steel samples as well as the increase in manganese content (compared to Fe–18Cr–12Mn–0·6N counterpart) also deteriorates the corrosion resistance.

These results could be taken as proof of some known results in the literature about the influence of manganese in other corrosion media. The added γ stabiliser manganese deteriorates the general corrosion resistance of different ferroalloys in non-oxidising acid environments, but does not have an effect on the passive current and the width of the passive region.^22–26 However, the latter is true only for environments in which local corrosion does not take place. It is well known that the manganese non-metallic inclusions are anodes in respect to the steel matrix. Thus, in the presence of active ions, the increase in manganese content above 1% results in facilitation of pit formation processes and hampers repassivation ones.²⁷ Still, taking into consideration the low solubility of manganese phosphates, it was interesting to check the corrosion behaviour of high manganese stainless steels regarding its pitting corrosion resistance. One would expect that the manganese corrosion products could impede the pits’ growth in acid phosphate solutions, although at low pH value. Unfortunately, the received results refute these expectations. The steel with the highest manganese content, Fe–14Cr–15Mn–0·2N, is the most susceptible to pitting corrosion in 5 and 40%H₃PO₄ solutions. For the above mentioned steel, the most negative values of E _pit and E _rp were registered, as well as pit development at lowest chloride concentration compared with other investigated steels. The narrowness of the passive region is more likely to be because of lower Cr and N content than of the higher content of Mn. It is known that the nitrogen improves the general corrosion resistance of stainless steels in chloride environments acidified by sulphuric acid, but it is almost inactive when the environment is of neutral and alkaline nature.^28–30 Moreover, nitrogen is shown to stabilise the passive state by facilitating the achievement of the passive state, decreasing the passive current and increasing the pitting corrosion resistance.^29,31–33

Thus, it can be established that the steel with higher nitrogen and chromium contents exhibits better corrosion resistance than the other manganese steel, but lower than the nickel steel. The lower corrosion resistance of Fe–18Cr–12Mn–0·6N steel compared to that demonstrated by the chromium–nickel samples indicates that, despite a positive effect of its higher nitrogen content (0·61 wt-%), it is not enough to compensate the presence of manganese and the absence of nickel in the composition.

Microscopic study after electrochemical corrosion tests

The addition of chlorides above a concentration of 0·04M to 5%H₃PO₄ leads to local corrosion attacks of the surface, for all the three steels tested. With samples of Fe–14Cr–15Mn–0·2N, for instance, local breakdowns of the passive layer were observed at chloride concentration higher than 0·005M, although pitting corrosion development was not registered by the polarisation method. The most probable explanation for this behaviour is related to defects owing to dissolution of anodic non-metallic inclusions.³⁴ Similar defects in the passive films, to smaller extent, were observed with Fe–18Cr–12Mn–0·6N and Fe–18Cr–9Ni steels.

With increase in the chloride content in the solution, the number of pits on Fe–18Cr–12Mn–0·6N steel surface increases. The pit sizes decrease, and a general corrosion attack to the entire surface not affected by the pitting process starts. In 5%H₃PO₄ with chloride content of 0·05M, for instance, the number of pits observed was ∼10 cm^{− 2}, of ∼120 μm in diameter, while the surrounding surface remained passive and shining. In addition, the test with 1·0M NaCl revealed an increase in the number of pits to several dozens: among them, micropits of dimensions below 30 μm predominated and were located mainly along the grain boundaries. Single open type large pits with deep irregular bottoms and mouths of ∼80 μm were also observed. Besides the pit formation, the corrosion attack affected the entire surface of the sample, which led to partial microetching of the structure. This effect probably occurs in the region of active dissolution where the kinetic controlled mechanism dominates. Additionally, the partial microetching of the structure was observed on the samples’ surfaces after OCP examination. It is known that the kinetic control leads to a dissimilar dissolution rate of different phases and microstructures.³⁵

Within the chloride concentration range from 0·05 to 0·1M, for instance, the pits on the chromium–nickel steel surfaces were of covered type and great depths. At higher chloride concentrations, the size and number of large pits for Fe–18Cr–9Ni steel grew, significantly exceeding those observed on chromium–manganese steel. The surface surrounding the pits was also subject of a general attack, but considerably weaker than that observed onto the chromium–manganese steels. The increased number of macropits as well as the feeble attack on the unaffected pits’ surface could be attributed to the increased resistance of the passive layers obtained during the anode polarisation of the chromium–nickel steel.

Intensification of the general attack on the surface was observed when the phosphoric acid concentration increased up to 40%, in presence of chloride ions. It was pronounced particularly in Cl⁻ concentrations beyond 0·1M by selective dissolution of single grains (probably with more suitable for the dissolution crystal lattice orientation) (Fig. 9a). On Fe–18Cr–12Mn–0·6N steel, well visible pits were observed in chloride concentrations over 0·5M. These pits were of an average size of 80-100 μm and irregular shapes. For 1·0M NaCl concentration, the total surface dissolution was significant, and any traces caused by preliminary machining treatments of the samples were completely erased. The pits had irregular shapes and occupied a considerable part of the surface (Fig. 10). The surface of Fe–18Cr–9Ni steel was more weakly attacked up to a chloride concentration of 0·5M. In the presence of 1·0M NaCl, the surfaces attained microetched structures (Fig. 9).

a surface out of pits; b general view of surfaceSurface of Fe–18Cr–9Ni steel samples after polarisation test in 40%H₃PO₄+1·0M NaCl

a general view of surface; b pit bottomSurface of Fe–18Cr–12Mn–0·6N steel samples after polarisation test in 40%H₃PO₄+1·0M NaCl

At acid concentration of 75%H₃PO₄, pitting corrosion was not observed even with maximum chloride concentration of ∼0·5M. The corrosion attack was general, spotted and went deeper with increase in the aggressive ion concentration, i.e. from weak etching structure at 0·05M NaCl towards strong corrosion and formation of deep caves of uneven, stepwise bottom. A small number of the grains were strongly attacked by the corrosion.

In the solution of 40%H₃PO₄+1·0M NaCl and anodic polarisation of samples of Fe–18Cr–12Mn–0·6N and Fe–18Cr–9Ni steels beyond 0·8 V(SCE), a green coloured compound was precipitated into the solution. In the literature background, such compound is termed as green rust and is commonly reported when mild steels are studied in phosphate solutions containing chlorides.³⁶ According to the authors, the green rust has different composition depending on the composition of the steel and most of all of the solution. In the experiments performed with phosphate solutions, the green rust was observed only in the presence of chlorides.

X-ray photoelectron spectroscopy analysis of passive layers on Fe–18Cr–12Mn–0·6N

X-ray photoelectron spectroscopy analyses were carried out to see what was the chloride effect on the composition of the passive layers formed during the polarisation of Fe–18Cr–12Mn–0·6N steel samples [anodically polarised at +0·4 V(SCE)]. The tests lasted 2 h in 40%H₃PO₄ with and without addition of 0·1M NaCl. The data reveal that, at this anodic polarisation, the density of the passive current attains a minimum, which, in general, corresponds to most stable state of the passive layer (to some extent related to its maximal thickness). Table 2 summarises results obtained for the composition of the surface layers.

Table 2

Compositions of surface layers/at-%

	O	N	Fe	Cr	Mn	P	Cl
40%H₃PO₄+0·1M NaCl	69·1	7·5	2·1	6·0	0·6	12·8	1·9
40%H₃PO₄	72·9	7·5	3·4	8·0	0·7	7·5	…

Several oxidation states were determined for the elements Cr, Fe, N and O, as well as their type and percentage share of the peak total areas for the respective elements, as they are presented in Fig. 11 and Table 3. Phosphorus has only one characteristic peak of its bond with oxygen, and it might be considered that the entire analysed quantity is in the form of phosphate. Therefore, the ratio (%O–4%P)/4%P represents the relation of the total oxygen quantity in the oxide, hydroxide and ads.H₂O to the oxygen in the phosphate groups in the passive layer. Spectrum decomposition was not performed for the elements Mn and Cl because of their low quantities.

X-ray photoelectron spectroscopy spectra of Cr 2p, Fe 2p, O 1s and N 1s for Fe–18Cr–12Mn–0·6N SS anodically polarised at 0·4 V(SCE) for 2 h in 40%H₃PO₄ with and without NaCl

Table 3

Percentage proportions of basic forms of elements Cr, Fe, O and N in passive layers formed in 40%H₃PO₄ with and without NaCl

	Cr	Fe	N	O
	Cr⁰/Cr³⁺ox (Cr₂N)/Cr³⁺(hy+ )	Fe⁰/Fe²⁺	N/NH₃/	Me ox/OH⁻ or /H₂O
40%H₃PO₄+0·1M NaCl	22·4/34·5/43·1	38·3/61·7	0/82·3/17·7	1·8/66·3/31·9
40%H₃PO₄	14·5/36·7/48·8	56·8/43·2	26·4/67·0/6·6	1·9/89·0/9·1

The data summarised in Table 2 reveal that the main differences in the compositions of the surface layers formed in 40%H₃PO₄, as results of the chloride action, are in the high content of phosphates and insertion of chlorine into the passive layers. The ratio (%O–4%P)/4%P considerably decreases from 1·47 to 0·35 in the presence of chlorides. Further, the amount of the water absorbed by the surface layers increases. Hence, in the solution of 40%H₃PO₄+0·1M NaCl, the main quantity of O is bonded to the phosphorus and the amounts of the metal oxides reduce while those of the phosphates increase.

The spectrum of the nitrogen at 397·8 eV corresponding to nitrogen in atomic state^31,32 or as a nitride^30,33 is missing when the surfaces were treated in 40% acid with chlorides. Despite this, the total amount of the nitrogen remains unchanged owing to the increase in the third peak at 402·2 eV, corresponding to . In this context, Vanini et al.,³² who reported that the bonds with energies of 400·2 and 402·1 eV are related to N–H ( is the suggested composition), may be cited. Moreover, they state that these bounds are localised only on the external layer of the passive film. According to Fu et al.,³⁰ particles NH₃ and in the passive layers are as a result of the reaction of nitride or N in the steel with the solution. The absence of a peak at 397·8 eV can be explained by the increased thickness of the passive layer and consequently the impossibility of the analysing beam to reach the metal/oxide interface, i.e. the location where there is a large accumulation of nitrogen on the metal surface.

The increased ratio of Cr/Fe, and the augmented amount of H₂O and , when nitrogen is missing in either atomic or nitride states when tests were carried in 40%H₃PO₄+0·1M NaCl, reveals that the passive layers are of increased thickness. It might be suggested that the anodic polarisation boost the chloride ions to build an additional external phosphate layer with increased amounts of water, chlorides and ammonia.

Conclusion

The presence of chlorides in phosphoric acid solutions of various concentrations affects unfavourably the general corrosion resistance as well as the formation and stability of the passive layers of chromium–manganese–nitrogen (Fe–18Cr–12Mn–0·6N, Fe–14Cr–15Mn–0·2N) and chromium–nickel (Fe–18Cr–9Ni) stainless steels. At applied anodic polarisation, the chlorides provoke pitting corrosion in solutions of 5 and 40%H₃PO₄, while in 75% acid, the irregular polishing effect was observed.

The composition of the stainless steels affects strongly the corrosion behaviour in the phosphoric acid solutions containing chloride ions. The replacement of nickel by 12% manganese and 0·6% nitrogen in Fe–18Cr–12Mn–0·6N steel reduces the general corrosion resistance and hinders the attainment of anodic passivity state. The passive region remains almost unaffected. The lower chromium and nitrogen content of Fe–14Cr–15Mn–0·2N steel compared to that in Fe–18Cr–12Mn–0·6N steel deteriorates further all corrosion characteristics: this is manifested by higher values of the corrosion rate, critical and passive currents as well as by increased pit nucleation at lower degree of anodic polarisation.

In general, it can be stated that the usage of the three examined steels in phosphoric acid solutions with concentration above 5% containing chloride over 0·01M is not acceptable because of the great uniform corrosion rate and pitting corrosion development. Implementation of anodic protection in these environments is inadmissible because of high risk of localisation of corrosion attack.

References

Rashev

T. V.

: ‘High nitrogen steels, metallurgy under pressure’ Publishing House of the Bulgarian Academy of Science ‘Prof. Marin Drinov’, Sofia; 1995.

Raj

and Mudali

U. K.

: ‘High nitrogen steels and stainless steels, manufacturing, properties and applications’; 2005

Narosa

New Delhi.

Abdel-Aal

M. S.

and Hermas

A. A.

: ‘Anodic polarization behaviour of cast iron alloys in deaerated concentrated H₃PO₄ solutions containing Cl⁻ and F⁻ ions’ J. Appl. Electrochem., 2000 30 339–345. doi: 10.1023/A:1003903806668

Iken

Basseguy

Guenbour

and Ben Bachir

: ‘Classic and local analysis of corrosion behaviour of graphite and stainless steel in polluted phosphoric acid’ Electrochim. Acta 2007 52 2580–2587. doi: 10.1016/j.electacta.2006.09.013

Prabakaran

and Rajeswari

: ‘Electrochemical, SEM and XPS investigations on phosphoric acid treated surgical grade type 316L SS for biomedical applications’ J. Appl. Electrochem., 2009 39 887–897. doi: 10.1007/s10800-008-9738-5

El Dahan

H. A.

: ‘Pitting corrosion inhibition of 316 stainless steel in phosphoric acid–chloride solutions. Part I: potentiodynamic and potentiostatic polarization studies’ J. Mater. Sci., 1999 34 851–857. doi: 10.1023/A:1004545501971

El Dahan

H. A.

: ‘Pitting corrosion inhibition of 316 stainless steel in phosphoric acid–chloride solutions. Part II: AES investigation’ J. Mater. Sci., 1999 34 859–868. doi: 10.1023/A:1004597518809

Guenbour

Bui

Faucheu

Segui

Ben Bachir

and Dabosi

: ‘The electrical properties of passive films formed on stainless steels in phosphoric acids’ Corros. Sci., 1990 30 189–199. doi: 10.1016/0010-938X(90)90073-E

Zhang

T. C.

Jiang

X. X.

and Li

S. Z.

: ‘Acceleration of corrosive wear of duplex stainless steel by chloride in 69% H₃PO₄ solution’ Wear 1996 199 253–259. doi: 10.1016/0043-1648(96)06982-7

10.

Escrivà-Cerdán

Blasco-Tamarit

García-García

D. M.

García-Antón

and Guenbour

: ‘Passivation behaviour of alloy 31 (UNS N08031) in polluted phosphoric acid at different temperatures’ Corros. Sci., 2012 56 114–122. doi: 10.1016/j.corsci.2011.11.014

11.

BenSalah

Sabot

Triki

Dhouibi

Refait

Ph.

and Jeannin

: ‘Passivity of Sanicro28 (UNS N-08028) stainless steel in polluted phosphoric acid at different temperatures studied by electrochemical impedance spectroscopy and Mott–Schottky analysis’ Corros. Sci., 2014 86 61–70. doi: 10.1016/j.corsci.2014.04.056

12.

Idrissi

: ‘Corrosion behaviour of two stainless steels and a chromium cast iron in phosphoric acid contaminated with S^{2 −} ions’ J. Mater. Sci., 1996 31 4599–4602. doi: 10.1007/BF00366357

13.

El Hajjaji

Aries

Audouard

J.-P.

and Dabosi

: ‘The influence of alloying elements on the corrosion resistance of stainless steels in phosphoric acid medium polluted by sulphide ions’ Corros. Sci., 1995 37 927–939. doi: 10.1016/0010-938X(95)00005-5

14.

Nassif

: ‘Influence of impurities in phosphoric acid on the corrosion resistance of some commercially produced alloys’ Surf. Technol., 1985 26 189–198. doi: 10.1016/0376-4583(85)90115-3

15.

Zuo

Wang

Zhao

and Xiong

: ‘The effects of some anions on metastable pitting of 316 stainless steel’ Corros. Sci., 2002 44 13–24. doi: 10.1016/S0010-938X(01)00031-2

16.

Refaey

S. A. M.

Abd El-Rehim

S. S.

Taha

Saleh

M. B.

and Ahmed

R. A.

: ‘Inhibition of chloride localized corrosion of mild steel by

and

anions’ Appl. Surf. Sci., 2000 158 190–196. doi: 10.1016/S0169-4332(00)00016-7

17.

Reffass

Sabot

Jeannin

Berziou

and Refait

Ph.

: ‘Effects of phosphate species on localised corrosion of steel in NaHCO₃+NaCl electrolytes’ Electrochim. Acta 2009 54 4389–4396. doi: 10.1016/j.electacta.2009.03.014

18.

Tang

Zhao

Mao

and Zuo

: ‘The electrochemical characteristics of pitting for two steels in phosphate buffer solution with chloride’ Mater. Chem. Phys., 2009 116 484–488. doi: 10.1016/j.matchemphys.2009.04.019

19.

Lakatos-Varsányi

Falkenberg

and Olefjord

: ‘The influence of phosphate on repassivation of 304 stainless steel in neutral chloride solution’ Electrochim. Acta 1998 43 187–197. doi: 10.1016/S0013-4686(97)00224-7

20.

Urretabizkaya

Pallotta

C. D.

De Cristofaro

Salvarezza

R. C.

and Arvia

A. J.

: ‘Changes in the composition of the passive layer and pitting corrosion of stainless steel in phosphate-borate buffer containing chloride ions’ Electrochim. Acta 1988 33 1645–1651. doi: 10.1016/0013-4686(88)80237-8

21.

Refaey

S. A. M.

: ‘Inhibition of steel pitting corrosion in HCl by some inorganic anions’ Appl. Surf. Sci., 2005 240 396–404. doi: 10.1016/j.apsusc.2004.07.014

22.

Tzaneva

B. R.

Loukaycheva

M. H.

and Fachikov

L. B.

: ‘Effect of phosphoric acid concentration on corrosion of Cr–Mn–N and Cr–Ni stainless steels’ Corros. Eng. Sci. Technol., 2014 49 219–227. doi: 10.1179/1743278213Y.0000000125

23.

Qiao

Y. X.

Zheng

Y. G.

Okafor

P. C.

and Ke

: ‘Electrochemical behaviour of high nitrogen bearing stainless steel in acidic chloride solution: effects of oxygen, acid concentration and surface roughness’ Electrochim. Acta 2009 54 2298–2304. doi: 10.1016/j.electacta.2008.10.038

24.

Vasudevan

Raja

V. S.

Seshan

and Chattopadhyay

: ‘Study of influence of manganese additions on electrochemical corrosion behaviour of austenitic spheroidal graphite iron’ Br. Corros. J., 1986 21, (2), 87–93. doi: 10.1179/000705986798272280

25.

Kemp

van Bennekom

and Robinson

F. P. A.

: ‘Evaluation of the corrosion and mechanical properties of a range of experimental Cr–Mn stainless steels’ Mater. Sci. Eng. A 1995 199 183–194. doi: 10.1016/0921-5093(94)09694-5

26.

Tzaneva

B. R.

Fachikov

L. B.

and Raichev

R. G.

: ‘Pitting corrosion of Cr–Mn–N steel in sulphuric acid media’ J. Appl. Electrochem., 2006 36 347–353. doi: 10.1007/s10800-005-9070-2

27.

Mitrovic-Scepanovic

and Brigham

R. J.

: ‘Effect of manganese on pitting properties of type 301L stainless steel’ Corrosion 1996 52 23–27. doi: 10.5006/1.3292091

28.

Muthupandi

Srinivasan

P. B.

Seshadri

S. K.

and Sundaresan

: ‘Corrosion behaviour of duplex stainless steel weld metals with nitrogen additions’ Corros. Eng. Sci. Technol., 2003 38 303–308. doi: 10.1179/147842203225008895

29.

Lothongkum

Wongpanya

Morito

Furuhara

and Maki

: ‘Effect of nitrogen on corrosion behavior of 28Cr–7Ni duplex and microduplex stainless steels in air-saturated 3·5 wt% NaCl solution’ Corros. Sci., 2006 48 137–153. doi: 10.1016/j.corsci.2004.11.017

30.

Han

E.-H.

Yang

and Jiang

: ‘Effects of nitrogen on the passivation of nickel-free high nitrogen and manganese stainless steels in acidic chloride solutions’ Electrochim. Acta 2009 54 4005–4014. doi: 10.1016/j.electacta.2009.02.024

31.

Olefjord

and Wegrelius

: ‘The influence of nitrogen on the passivation of stainless steels’ Corros. Sci., 1996 38 1203–1220. doi: 10.1016/0010-938X(96)00018-2

32.

Vanini

A. S.

Audouard

J.-P.

and Marcus

: ‘The role of nitrogen in the passivity of austenitic stainless steels’ Corros. Sci., 1994 36 1825–1834. doi: 10.1016/0010-938X(94)90021-3

33.

Jiang

Yang

Cao

and Zhang

: ‘Pitting corrosion and crevice corrosion behaviors of high nitrogen austenitic stainless steels’ Int. J. Miner. Met. Mater., 2009 16 517–524. doi: 10.1016/S1674-4799(09)60090-X

34.

Burstein

G. T.

Liu

Souto

R. M.

and Vines

S. P.

: ‘Origins of pitting corrosion’ Corros. Eng. Sci. Technol., 2004 39 25–30. doi: 10.1179/147842204225016859

35.

Chen

S. C.

G. C.

and Huang

C. A.

: ‘The electrochemical polishing behaviour of porous austenitic stainless steel (AISI 316L) in phosphoric–sulphuric mixed acid’ Surf. Coat. Technol., 2005 200 2065–2071. doi: 10.1016/j.surfcoat.2005.06.008

36.

Simard

Odziemkowski

Irish

D. E.

Brossard

and Menard

: ‘In situ micro-Raman spectroscopy to investigate pitting corrosion product of 1024 mild steel in phosphate and bicarbonate solutions containing chloride and sulfate ions’ J. Appl. Electrochem., 2001 31 913–920. doi: 10.1023/A:1017517618191