Evaluation of the Corrosion Behavior of Low-Temperature Nitrided AISI 316L Austenitic Stainless Steel ^†

Abstract

Nitriding of austenitic stainless steels at low temperatures hinders the precipitation of chromium nitrides and causes the formation of a supersaturated solid solution of nitrogen atoms in the austenite lattice, known as expanded austenite. In this study, the corrosion behavior of low-temperature nitrided AISI 316L is investigated in a NaCl solution using different electrochemical techniques, electrochemical impedance spectroscopy, cyclic potentiodynamic polarization and galvanostatic tests, in order to assess the effect of test conditions. The nitrided layer has an enhanced resistance to localized corrosion, but its ability to repassivate depends on the damage extent caused by the different tests.

1. Introduction

Austenitic stainless steels are used in many industrial fields, thanks to their good corrosion resistance in many environments, the ease of formability and weldability, and the possibility to employ them even at cryogenic temperatures [1,2]. However, their low hardness and poor tribological properties may limit their application. Surface engineering techniques have become a valuable tool for prolonging the life of austenitic stainless steel components. Among the different techniques, low-temperature nitriding is particularly interesting [3,4,5]. When nitriding is performed at the temperatures usually employed for low-alloy steels (500–550 °C), chromium (Cr) nitrides are easily formed, and thus a decrease in corrosion resistance is observed, since the Cr-depleted matrix cannot maintain a protective passive film. On the contrary, when nitriding is performed at temperatures lower than 450 °C, nitrogen (N) atoms can freely diffuse, while the diffusion of substitutional atoms, such as Cr, is significantly slowed down, and therefore the formation of Cr nitrides tends to be hindered. Using these treatment conditions, a supersaturated solid solution of N in the face-centered cubic (fcc) austenite lattice is formed, known as expanded austenite or S-phase, and an enhancement of surface hardness can be obtained. The corrosion resistance of the nitrided layer depends on its microstructure and phase composition, as well as on the environment characteristics [5]. The testing conditions can also play a role in its corrosion behavior. As an example, conflicting results have been obtained on the capability of nitrided austenitic stainless steels to repassivate when cyclic potentiodynamic polarization tests have been performed [6,7,8,9].

The aim of the present study is to evaluate the corrosion behavior of low-temperature nitrided AISI 316L austenitic stainless steel in a NaCl solution using different electrochemical techniques, electrochemical impedance spectroscopy (EIS), cyclic potentiodynamic polarization and galvanostatic tests, paying special attention to the repassivation capability.

2. Materials and Methods

AISI 316L specimens (40 × 17 × 0.7 mm) were grinded and polished up to 6 µm diamond paste. The nitriding treatments were carried out using the d.c. glow-discharge process, as previously described [10]. The treatments were carried out using a laboratory plasma equipment, where the sample holder was connected to the cathode of a d.c. power supply (direct nitriding). After a sputtering step, nitriding was performed at 380 °C, 130 Pa, for 5 h using a gas mixture of 80 vol. % N₂ + 20 vol. % H₂. The voltage drop was 253 ± 5 V and the current density was 0.90 ± 0.05 mA cm⁻². With the used test conditions, the edge effect on the specimens was completely negligible in the regions used for the subsequent analyses of the nitrided samples.

The untreated and treated specimens were examined by means of light microscopy technique. The cross-section microstructure was delineated using acetic glyceregia etchant (3 mL HCl, 2 mL HNO₃, 2 mL CH₃COOH, 1 drop of glycerol). Phase analysis was carried out using X-ray diffraction (XRD) in Bragg–Brentano configuration (Cu Kα radiation, λ = 1.5406 Å; scan speed: 0.1° s⁻¹).

Knoop microhardness measurements were performed on the surface of the specimens using a load of 25 g_f.

Corrosion behavior was assessed in a 5 wt.% NaCl aerated solution at room temperature without stirring. A three-electrode electrochemical flat cell was used, having a platinum grid as counterelectrode and an Ag/AgCl reference electrode (3.5 M KCl). The sample surface area exposed to the solution was 1 cm². The electrochemical tests were performed after a 20 h delay. At least three tests for each sample type were carried out to assess the result. Electrochemical Impedance Spectroscopy (EIS) measurements were performed at the Open Circuit Potential (OCP) in the frequency range 10 kHz–12 mHz, using an ac amplitude (peak-to-peak) of 2.5 mV. EIS spectra were modeled using non-linear least square analysis software (Gamry Echem Analyst, vers. 7.06). Galvanostatic tests were performed imposing a constant anodic current of 100 µA cm⁻² and the potential variation was recorded for 3000 s. Cyclic potentiodynamic polarization tests were very sensitive to the scan rate, in particular regarding repassivation potential values. As an example, the polarization curves of untreated specimens tested using forward and backward scan rates of 0.3 or 1 mV s⁻¹ are depicted in Figure 1. Taking into account the results of preliminary tests, polarization was carried out with a forward scan rate of 0.3 mV s⁻¹ up to 100 µA cm⁻², and then the backward scan was performed at 1 mV s⁻¹.

3. Results and Discussion

The formation of the modified surface layer causes local plastic deformations, which are observable at the surface (Figure 2a). The analysis of the cross-section shows that the nitrided layer is continuous and fairly homogeneous, highly resistant to chemical etching, and it has a two-layer microstructure (Figure 2b). In the outer layer, expanded austenite, γ_N, is present, together with a small amount of a hexagonal close packed (hcp) phase, which can be regarded as an N-induced ε’-martensite, ε_N’ (Figure 2c). It is hypothesized that ε_N’ forms from γ_N as a consequence of a fcc-to-hcp martensitic transformation, due to the plastic deformations caused by the huge N solubilization [11,12]. In the inner part of the modified layer, a solid solution of interstitial atoms in austenite, γ(N, C), is detected. The outer layer has a thickness of 6.8 ± 0.4 µm, the inner layer of 1.0 ± 0.1 µm.

The nitrided layer allows for a significant increase in surface hardness, 1473 ± 108 kg_f mm⁻², almost 5.5 times the microhardness of the untreated alloy (268 ± 4 kgf mm⁻²).

The typical EIS spectra of untreated and nitrided samples are depicted in Figure 3 in the form of Bode (a) and Nyquist (b) plots. The nitrided samples have higher impedance data and tend to form a wider semicircle in the Nyquist plot, suggesting a higher resistance to general corrosion. A physical picture of the corrosion behavior is obtained modeling the experimental data with the electrical equivalent circuit (EEC) depicted in the inset A. Constant phase elements (CPEs) are used instead of pure capacitance in parallel with the resistance (R) elements. Taking into account the asymmetry in the phase angle plot, a model with two time constant elements is employed. The high frequency time constant (R_ct//CPE_dl) is related to the charging/discharging processes taking place at the electrode–electrolyte interface, while the low frequency time constant (R_o//CPE_o) is related to the charge transfer and mass transport occurring in the oxide layer [13,14,15]. The EEC parameter values are reported in

Table 1. Best fitting EEC parameter values for EIS spectra of samples untreated and nitrided, tested at the respective OCP (model: inset A in Figure 3b).

Sample Type	Rs (Ω cm2)	Rct (MΩ cm2)	CPEdl × 105 (Ω−1 sn cm−2)	ndl	Ro (MΩ cm2)	CPEo × 105 (Ω−1 sn cm−2)	no	Rtot (MΩ cm2)
untreated	5.4 ± 0.5	1.1 ± 0.3	2.9 ± 0.3	0.96 ± 0.02	1.7 ± 0.5	1.6 ± 0.5	1.00 ± 0.2	2.8 ± 0.8
nitrided	7.9 ± 0.5	4.7 ± 0.5	1.7 ± 0.3	0.94 ± 0.02	113 ± 30	0.13 ± 0.03	0.9 ± 0.2	117 ± 30

. An increase in the resistance components and a decrease in the CPE terms are registered for the nitrided samples, if compared to those of the untreated alloy. The sum of R_ct and R_o resistance terms, R_tot, can be considered a measure of the resistance to general corrosion, and hence a significant enhancement can be supposed for the nitrided specimens.

Cyclic potentiodynamic polarization tests only give a partial confirmation of the EIS results. As depicted in Figure 4a, both treated and untreated samples have polarization curves typical of a passive-transpassive behavior, but the nitrided samples have higher corrosion and pitting potential values and lower anodic current density in the passive branch, if compared to those of the untreated steel. A few slight increases and decreases in anodic current density values are registered in the passive branch of the nitrided samples, suggesting the formation of metastable pits. The oxide film is locally broken and metal dissolution occurs until a fast repassivation. However, when a 100 µA cm⁻² anodic current density is reached, the two sample types have a different behavior. For the untreated steel, the current density tends to increase even if the potential is decreasing, due to the autocatalytic nature of localized corrosion phenomena [16], but a significant decrease in the current density is observed at about +194 mV (Ag/AgCl), suggesting that repassivation occurs. After the tests, a few pits are present on the surface (Figure 4b). For the nitrided samples, the anodic current density values remain around ~100 µA cm⁻² for potential values even lower than corrosion potential, suggesting that no repassivation occurs. After the test, many wide colored regions are observable on the surface, due to corrosion products (Figure 4c).

The galvanostatic tests give a different picture, as shown in Figure 5. The untreated specimens have the typical trend of a passive material [15,17], with a fast increase in potential up to a maximum, which corresponds to the beginning of surface activation, followed by a decrease, corresponding to the propagation of localized corrosion phenomena, with potential fluctuation due to the breakdown/repair of the passive film. Then the potential tends to reach a stationary value, which can be regarded as the minimum potential, below which localized corrosion phenomena do not occur [15,18]. This potential value is higher than the corrosion potential, and therefore the occurrence of repassivation may be hypothesized. After the test, a few large pits are observable on the surface. For the nitrided samples, a high potential value (about +1155 mV (Ag/AgCl)) is quickly attained, and then the potential values remain high with some fluctuations, which can be ascribed to damages to the nitride layer and repassivation phenomena. A significant potential decrease is observed near the end of the test, but the attained potential value is higher than the corrosion potential of both nitrided and untreated specimens, suggesting that, with the used test conditions, repassivation can occur. After the test, the surface of the samples is slightly discolored, due to transpassive dissolution of the oxide film, and a few small pits are observable.

The results of the electrochemical tests suggest that the nitrided samples have an enhanced resistance to the occurrence of localized corrosion phenomena, but their ability to repassivate depends on the test conditions (potentiodynamic/galvanostatic) and hence on the extent of the corrosion damage. A fairly large corrosion damage, as that produced by the test conditions used in the cyclic potentiodynamic polarization tests, seems to hinder the repassivation.

4. Conclusions

Low-temperature nitriding allows for producing, on AISI 316L austenitic stainless steel, a modified surface layer mainly consisting of expanded austenite, which improves surface hardness and has an enhanced resistance to localized corrosion phenomena in NaCl solution. However, the ability of the nitrided layer to repassivate is particularly sensitive to the extent of corrosion damage. When a small surface damage is present, as that produced by galvanostatic tests, the potential values attained during the test suggest that repassivation can occur. On the contrary, a fairly large corrosion damage, such as that caused by cyclic potentiodynamic polarization, tends to hinder repassivation.