Investigation on the Relationship between the Surface Chemistry and the Corrosion Resistance of Electrochemically Nitrided AISI 304 Stainless Steel

The relationship between the surface chemistry and the corrosion resistance of electrochemically nitrided AISI 304 stainless steel samples has been investigated. The nitriding treatment was carried out in HNO3 0.1M and HNO3 0.1M + KNO3 0.5M at room temperature. Samples were subjected to the nitriding procedure for 30 minutes under a cathodic potential of -0.7 VAg/AgCl. The chemical composition of the nitrided layers was assessed by X-ray photoelectron spectroscopy (XPS). Depth profiles of the main elements present in the nitrided layers were also obtained by XPS by etching them with argon ions. The corrosion behavior of the nitrided samples was evaluated by potentiodynamic polarization.The results showed that the nitrided layers consisted of a mixture of chromium nitrides, chromium oxides, iron oxides/oxyhydroxides, and nickel oxide. The best corrosion resistance was obtained by electrochemical nitriding in the HNO3 0.1M + KNO3 0.5M solution. This result could be correlated with the composition and thickness of the nitrided layer.


Introduction
Stainless steels are prone to localized corrosion in contact with chloride-containing solutions [1][2][3].The stability of the naturally formed passive film accounts for its corrosion resistance in a variety of environments [4][5][6].One promising way of increasing the reliability of stainless steels is to increase passive film stability [7,8].Nitrogen has been shown to improve the pitting and crevice corrosion resistances of stainless steels [9].It is reported to enable the widening of passive range, decrease of passive current density, thus leading to positive effects on the resistance to stress corrosion cracking and intergranular corrosion [10,11].The mechanisms through which the beneficial effects of nitrogen take place are yet unclear.Several hypotheses have been proposed such as nitrogen incorporation into the passive film, suppressing anodic dissolution, the formation of nitrate ions with repulsive action against Cl − , or neutralization of the acidity inside corrosion pits by the formation of NH 4 + ions, thus resulting in repassivation [12][13][14][15][16].
Nitrogen incorporation into stainless steels can be achieved by different methods.Alloying is employed for highnitrogen austenitic stainless steels but it is limited by the low nitrogen solubility in the iron matrix [17].Nitriding plays a central role in this scenario, giving more versatility to the nitrogen content that can be added to austenitic stainless steels [18].Plasma nitriding has gained increased technological relevance as a large-scale processing method to achieve nitrogen-rich stainless steel surfaces [19][20][21].Recently, electrochemical nitriding has emerged as an alternative route to incorporate nitrogen into the surface of stainless steels [22].Its main advantage when compared to other nitriding treatments relies on the possibility of conducting the process at room temperature.High temperature nitriding processes are typically undertaken above 500 ∘ C and low temperature ones take place at temperatures as low as 350 ∘ C [23,24].In this respect, electrochemical nitriding is regarded as an economic route.Pioneering works by Willenbruch et al. [25] and Kim et al. [26] investigated the nitriding mechanism of stainless steels in solutions consisting of HCl and NaNO 3 .Nitrate ions would adsorb on the metallic surface, being later reduced to NH 4 + and metal nitride.More recently, some authors have proposed the use of electrochemical nitriding to attain high corrosion resistance and conductivity for stainless steels employed as bipolar plates for polymer electrolyte membrane fuel cells (PEMFCs) [27][28][29].The composition of the nitrided layer plays a major role in the corrosion behavior of the treated surface.Wang et al. [29] observed the formation of nitrogen incorporated oxide (oxi-nitrides) in the surface of electrochemically nitrided AISI 446 stainless steels.

International Journal of Corrosion
In spite of the existing literature, similar information is not found for the AISI 304 austenitic stainless steel in conventional chloride-containing solutions.In this work, AISI 304 stainless steel specimens were electrochemically nitrided in different nitrate-containing electrolytes.The relationship between the chemical composition of the nitrided layer and the corrosion resistance of the treated surfaces was investigated.Surface analysis was performed by X-ray photoelectron spectroscopy (XPS).The corrosion behavior was evaluated by electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization.

Experimental
2.1.Material and Electrochemical Nitriding Procedure.Coldrolled AISI 304 stainless steel plate with a thickness of 4 mm was used as the substrate for electrochemical nitriding.Specimens for the surface treatment and subsequent characterization were square-shaped pieces with approximate dimensions of 2.5 × 2.5 cm.
Prior to electrochemical nitriding, the specimens were subjected to mechanical grinding using silicon carbide waterproof papers up to grit 600, being subsequently washed with deionized water and dried in a warm air stream provided by a conventional heat gun.Next, the specimens were electrochemically nitrided in different solutions, as described in Table 1.The treatment was performed using a conventional three-electrode cell setup with a platinum wire as the auxiliary electrode, Ag/AgCl as the reference electrode, and the AISI 304 specimen as the working electrode.Nitriding was performed by cathodic polarization at -0.7 V Ag/AgCl .The treatment time was 30 minutes.Two different electrolytes were employed as shown in Table 1.All experiments were carried out at room temperature using an Ivium n-Stat potentiostat/galvanostat.After nitriding, the specimens were thoroughly washed with deionized water and dried in a warm air stream provided by a heat gun.

Electrochemical Characterization.
The electrochemical behavior of the nitrided samples was assessed using the same three-electrode cell configuration employed for the nitriding treatment.The tests were carried out in 3.5 wt.% NaCl solution at room temperature.This electrolyte was selected as a common medium to evaluate the corrosion properties of metallic materials comparatively.Initially, the open circuit potential was monitored for 1 h.Next, the samples were subjected to potentiodynamic polarization at a sweep rate of 1 mV.s −1 from -300 mV versus the OCP up to +1.0 V Ag/AgCl .The tests were conducted in triplicate.After polarization, the corrosion morphology was examined by confocal laser scanning microscopy (Olympus, LEXT OLS4100).

Surface Analysis.
XPS analyses of the as-nitrided layers were carried out in a ThermoVG K-alpha + spectrometer operating with Al-k X-ray source.The pressure in the analysis chamber was 10 −6 Pa.Binding energy scale was calibrated using the C1s peak at 284.8 eV due to the presence of surface contamination (adventitious carbon).Data analysis was performed by Avantage software.Peak fitting was carried out by a weighted sum of Lorentzian and Gaussian curves, after background subtraction by the Smart algorithm.Depth profiles of Cr, Fe, Ni, O, and N were obtained by sputtering the nitrided surface with argon ions.Each sputtering cycle consisted of 15 s of exposure to the argon ions flux at 3 keV.

X-Ray Photoelectron Spectroscopy (XPS)
. XPS high resolution spectra for the Cr2p 3/2 core levels of the N and NK samples are shown in Figure 1.The spectrum for the untreated material is also shown for comparison (Figure 1(c)).The presence of CrN and Cr 2 N was observed as indicated by the peak at approximately 574 eV for the nitrided samples (Figures 1(a) and 1(b)) whereas it was not observed for the untreated steel (Figure 1(c)).The other two peaks are related to Cr-O bonding typical of Cr 2 O 3 and CrO 3 compounds [22].The main species in the surface film are oxides, but chromium nitrides were unequivocally identified too, revealing the incorporation of nitrogen during the electrochemical nitriding procedure for both nitrided samples.Table 2 shows the atomic fraction of each species in the surface film with regard to the Cr2p 3/2 core level.It is interesting to note that chromium nitrides were preferentially formed over sample NK, even considering the experimental error for XPS quantitative analysis (±10%) [30].The peak corresponding to metallic chromium (Cr 0 ) was only detected for the untreated sample.
The spectra for the N1s core levels of samples N and NK are shown in Figures 2(a) and 2(b), respectively.The spectra unequivocally confirm the formation of chromium nitride species as observed at 397.6 eV which is in good agreement with the literature [31,32].In addition, the peak at higher binding energies is ascribed to NH 3 groups that are present in smaller fractions when compared to chromium nitrides.The quantitative analysis of the fitted spectra is presented in Table 3.The presence of NH 3 is reported by other authors, being associated with the reaction of nitrides with adsorbed    water in the electrolyte [22].The data shown in Table 3 reveal a predominance of chromium nitride species over NH 3 groups for the either N or NK samples.The electrochemical nitriding procedure conducted at both solutions yielded a similar composition for the nitrided layer with respect to Cr and N species.
The composition of the surface layer was further characterized by evaluating the Fe2p 3/2 core levels.The spectra are shown in Figure 3.The fitting results are presented in Table 4.
Fitting results for the Fe2p 3/2 spectra indicated the presence of metallic iron (Fe 0 ) and a mixture of iron oxides and oxyhydroxides either for the untreated or nitrided samples.The binding energies are in good agreement with the literature [33][34][35].Metallic iron signal arises from the steel substrate [6].The results point that the surface of the untreated sample contains both iron and chromium oxides and hydroxides but nitrides are not present.The surface of the nitrided samples, in turn, contains chromium nitrides in addition to chromium and iron oxides/hydroxides.It is  interesting to note that the metallic chromium signal was absent for the nitrided layers but the metallic iron signal was clearly detected.Moreover, it was even more intense for the nitrided layers than on the untreated surface.Our results point to an enrichment of chromium nitrides and oxide species with the electrochemical nitriding procedure whereas the iron oxides and hydroxides were proportionally lowered with respect to the metallic iron signal.
The spectra for the Ni2p 3/2 core levels are shown in Figure 4.The main signal is referred to metallic nickel.Small fractions of NiO species are also detected at higher binding energies.
The strong Ni 0 signal is frequently reported for passive films formed on stainless steel surfaces, being associated with the lower oxidation susceptibility of nickel when compared to chromium and iron [36].The fitting results for the Ni2p 3/2 core levels are displayed in Table 5.The peak energies of Ni 0 and NiO species are in good agreement with the literature [37].It is interesting to note that the N sample presents a higher content of Ni0 when compared with the NK sample which, in turn, shows Ni 0 content compatible with that of the untreated steel.According to Dadfar et al. [6] the signal of the metallic substrate is an indication that the surface film is not homogeneous.Likewise, the stronger Ni 0 signal for the N sample suggests that its nitrided layer can be less homogeneous when compared to that of the NK sample.
The O1s spectra are shown in Figure 5. Fitting of these spectra showed two components for all samples, O 2− and OH − .These species are typical of stainless steel passive films and also for nitrided layers [6,38,39].The fitted data are displayed in Table 6.It is seen that the NK sample is more oxidized than the N sample, since the OH − species is present in higher contents.
By evaluating the XPS high resolution spectra, it was found that the nitrided layers consisted of a mixture of chromium nitrides, iron oxides/hydroxides, and nickel oxide.The main differences between N and NK samples are the lower NiO content of the N sample and higher OH − content of the NK sample.Additional analysis of the nitrided surfaces was based on XPS depth profiling.Depth profiles for Cr, Fe, O, N, and Ni are shown in Figure 6 for the N and NK samples.The nitrogen content was higher for the NK sample (Figure 6(b)) than for the N sample (Figure 6(a)).Nitrogen incorporation was favored for this sample as indicated by its higher atomic fraction, suggesting the mixed electrolyte was more effective as a nitrogen source for the stainless steel surface.
Yet, it is noteworthy that the oxygen atomic concentration was less affected by the sputtering cycles for the NK sample as shown in Figure 6(b).For the N sample, the passive film is more easily removed (Figure 6(a)).Furthermore, the nitrogen atomic fraction is lower for the N sample.As a consequence of the removal of the nitrided layer, the iron concentration shows an opposite trend with respect to the oxygen and nitrogen concentration, increasing sharply with the sputtering time.
The chromium content was higher for the NK sample, suggesting its passive film is enriched in chromium, thus favoring the corrosion resistance.The nickel concentration in the surface layer is much lower than that of chromium either for N or NK samples.(less cathodic), indicating the electrochemical reactions are more intense when the samples are treated in this electrolyte.The sequence of events during electrochemical nitriding is the reaction between metal and nitrate ions, leading to an adsorbed nitrate, followed by the reaction of this adsorbed species with H + ions, giving rise to metal nitrides.One additional reaction occurs between nitrate ions and H + , forming ammonium ion [40,41].The formation of chromium nitrides was unequivocally observed by the XPS results presented in Figure 1 as well NH 3 species (Figure 2).

Open Circuit Potential. The variation of the open circuit
potential with the immersion time in 3.5 wt.% NaCl solution at room temperature is shown in Figure 8 for the untreated and nitrided samples.The untreated sample presented the most negative potential values over the whole monitoring time, indicating a higher electrochemical activity when compared to the nitrided ones.There was an ascending trend of the open circuit potential in the beginning of the monitoring period for all samples, indicating that the passive film is likely to become thicker with time.The N sample presented the more anodic potential up to 2800 s when it presented a decreasing trend up to the end of the monitoring period, indicating that the passive film was no longer stable as at the first part of the test.The NK sample, in turn, presented a stable potential up to the end of the test, suggesting it is less prone to corrosion when compared to the N sample and the untreated substrate [42,43].The results suggest that the electrochemical nitriding treatment decreased the susceptibility of the 304 stainless steel substrate to corrosion and that the nitrided layer of the NK samples present the best stability in the electrolyte.

Potentiodynamic Polarization
Curves.Potentiodynamic polarization curves for the untreated and nitrided samples immersed in 3.5 wt.% NaCl solution at room temperature are shown in Figure 9.
The electrochemical parameters determined from the potentiodynamic polarization curves are displayed in Table 7.The corrosion potential (E corr ) of the NK sample is the noblest, being slightly more anodic than the value for the N sample.The untreated material, in turn, in spite of presenting the lowest value of E corr presented the highest pitting potential (E pit ), suggesting that it is more resistant to the onset of pitting corrosion.Notwithstanding, its passive current density (i pass ) was the highest, indicating that the corrosion kinetics is higher for the untreated surface.When the metallic surface  is in the passive state, the passive current density is a measure of the corrosion resistance of the material [44].The values of i pass were determined at the middle of the passive region, according to Ningshen et al. [45].Additionally, the protection efficiency (P%) of the electrochemical nitriding procedure was calculated based on equation (1), where i pass-t is the passive current density for the treated samples and i pass is the passive current density for the untreated sample.The values of protection efficiency were 33% for the N sample and 62% for the NK sample based on the mean values of i pass presented in Table 7.
The NK sample presented an intermediate pitting potential and the lowest value of i pass .Additionally, it is important to observe that the current spikes observed in the passive region of the untreated material are not found for the nitrided samples.These features are associated with metastable pitting, being related to the instability of the passive film and imminence of its breakdown [46,47].Although the susceptibility to pitting corrosion is related to the E pit , current spikes indicating metastable pitting are undesirable, since the passive film does not sustain its stability through the whole potential range.
The nitrided layer of the NK sample can be considered stable, since it does not show the formation of metastable pits and presents the lowest value of i pass .However, it is not immune to the onset of localized corrosion.Furthermore, even though the results displayed in Table 7 point to an improvement of the corrosion resistance after the electrochemical nitriding procedure, the standard deviations must be taken into account.In this respect, the differences of ipass do not point to a marked improvement, especially for the N condition.
In order to give further understanding about the protective efficiency of the nitrided surfaces, confocal laser scanning microscopy (CLSM) was employed to examine the corrosion morphology of the specimens subjected to potentiodynamic polarization.Representative pits and their corresponding transverse profiles are shown in Figures 10, 11, and 12 for the N, NK, and untreated samples, respectively.
It is interesting to note that pit growth was characterized by the formation of small pits surrounding bigger ones for the untreated sample, as observed in Figures 12(a  The results are displayed in Table 8.Pit diameter was lower for the nitrided samples whereas pit depth was similar the NK and the untreated material.Notwithstanding, the lowest diameter of the NK condition suggests its pit growth was slow, in accordance with its low i pass values shown in Table 7.These results can be correlated with those obtained by XPS.The relative poor corrosion resistance of the N sample when compared to NK would derive from its less thick nitrided layer.Depth profiles shown in Figure 6 clearly indicate that the nitrided layer formed on the NK sample is not as easily removed as that produced on the N samples, suggesting it is thinner and, therefore, less protective.This effect would be dominant over any difference of chemical composition.Indeed, the main composition differences were not observed for the chromium nitride compounds, but for the metallic nickel signal and for the oxygen species present on the sample surface.The higher metallic nickel signal of the N sample would indicate its less oxidized nature as well as its lower OH − content when compared to NK.These features would arise from the higher current densities experienced by the samples subjected to electrochemical nitriding in the HNO 3 0.1 M + KNO 3 0.5 M solution (NK samples), giving rise to a more protective nitrided layer.In this respect, the nitrided layer formed on NK samples, although it is more heterogeneous than the layer produced on the N samples (as shown by the XPS depth profiles presented in Figure 6), provided the best corrosion resistance due to a combination of its more oxidized state and higher thickness.
Our results point to a marginal improvement of the corrosion resistance for the N samples and a better protective ability for the NK condition.It is important to emphasize that the electrochemical nitriding treatment undertaken in the present work can be optimized to reach more significant performance against corrosion, exploring other concentrations of the chemical species or treatment times.The main contribution of this work is to firstly investigate the correlation between the surface chemistry and the corrosion resistance of electrochemically nitrided AISI 304 stainless steel samples.In this respect, it is the first step into a deeper understanding of the marked influence of surface chemistry and the possibility of controlling the corrosion resistance by modifying the electrolyte composition.

Conclusions
Electrochemical nitriding of AISI 304 stainless steel samples was successfully performed in HNO 3 0.1 M and HNO 3 0.1 M + KNO 3 0.5 M at room temperature by applying a cathodic potential to the steel samples.The nitrided layers consisted of a mixture of chromium nitrides with chromium oxides, iron oxides/oxyhydroxidesa, and nickel oxide.The samples nitrided in the HNO 3 0.1 M + KNO 3 0.5 M solution (NK) presented higher atomic concentrations of nitrogen species across the thickness of the nitrided layer, as shown by XPS depth profiles.The lowest values of passive current densities for the NK samples and its higher pitting potential indicated that the corrosion resistance of the nitrided samples was affected by the solution employed for electrochemical nitriding.NK samples presented an apparently thicker nitrided layer, although it grew more heterogeneously than the layer formed on the N samples, as indicated by the XPS depth profiles.The results show that it is possible to control the corrosion resistance of the nitrided layer, depending on the electrolyte employed for the nitriding treatment.

Figure 8 :Figure 9 :
Figure 8: Variation of the open circuit potential with the immersion time for the untreated and nitrided samples.Electrolyte: 3.5 wt.% NaCl solution at room temperature.

Figure 10 :
Figure 10: CLSM micrograph of a pit (a) and its corresponding transverse profile (b) for a representative N sample.

Figure 11 :
Figure 11: CLSM micrograph of a pit (a) and its corresponding transverse profile (b) for a representative NK sample.

Figure 12 :
Figure 12: CLSM micrographs for a representative untreated sample ((a) and (b)); (c) the transverse profile for the pit shown in (b).

Table 3 :
XPS fitting results for the N1s core level.

Table 6 :
XPS fitting results for the O1s core level.

Table 7 :
Electrochemical parameters determined from the potentiodynamic polarization curves.

Table 8 :
Pit dimensions determined by CLSM.