ISIJ International
Online ISSN : 1347-5460
Print ISSN : 0915-1559
ISSN-L : 0915-1559
Surface Treatment and Corrosion
Nitrogen Supersaturation Process in the AISI420 Martensitic Stainless Steels by Low Temperature Plasma Nitriding
Abdelrahman FarghaliTatsuhiko Aizawa
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2018 Volume 58 Issue 3 Pages 401-407

Details
Abstract

A high-density RF-DC plasma nitriding system was employed on AISI420-J2 martensitic stainless steel at 653 K, 673 K, and 693 K for 14.4 ks. Scanning electron microscope (SEM), energy dispersive spectroscopy (EDS), x-ray diffraction (XRD), and electron backscattered diffraction (EBSD) were utilized to make analysis and characterization of the nitrided layers. These layers with the thickness of 85 µm from the surface was mainly nitrogen super-saturated with formation of nitrides at the vicinity of surface. The nitrogen content depth profile was nearly constant by 10 at% except for the gradual decrease from the maximum content by 30 at% at the surface and for the decay toward the nitriding front end. The lattice expansion by the strain of 1.6% drove the phase transformation from the original martensite to austenite. High plastic straining following this elastic lattice expansion also caused the grain size refinement from the original size of 10 µm down to 0.15 µm.

1. Introduction

AISI420-J2 martensitic stainless steels have been widely utilized as a mold material for injection molding process. Their surface hardening and strengthening was keenly needed to prolong the life time of mold together with fine surface profiles. Various surface treatments were employed to improve the strength and surface hardness of stainless steels; e.g., the case hardening, the liquid and gas nitriding processes,1) and, the plasma nitriding related processes.2,3,4) In particular, many studies were reported on the effect of the nitriding parameters on the microstructure morphology as well as the wear and corrosion resistance of the nitrided surface.5,6) Among various parameters such as the alloying elements,7) the cathode setups,8) the gas mixtures,9) the nitriding times,10) and the nitriding temperatures,11,12) the holding temperature has the most significant impact on the nitriding processes as well as the mechanical and chemical properties of nitrided stainless steels.

In the plasma nitriding process above 723 K, the chromium nitrides (CrN) precipitate in the nitrided matrix; e.g., the surface hardness increased but with significant reduction in the corrosion resistance.13) On the other hand, no chromium nitrides precipitate in the plasma nitriding below 723 K.14,15,16) This low temperature plasma nitriding is characterized by the nitrogen supersaturation process to harden and strengthen the stainless steels by high nitrogen solute concentration and microstructure refinement without sacrificing their corrosion resistance.17,18) In addition, the expanded austenite formed in the AISI304 and AISI316 matrices19,20) and expanded martensite, in the AISI420 matrix.7,11,21) The expanded phases are supersaturated with nitrogen interstitials; e.g., much lattice expansion was observed up to 10%.22) These phases are considered to be metastable to drive the hardening process and to improve the corrosion resistance.23) In addition, the phase transformation takes place from martensite (α′) to austenite (γ) in the martensitic stainless steels.24) The above new features are all intrinsic to the nitrogen supersaturation process in the low temperature plasma nitriding.

In the present study, high density RF/DC plasma generation system is used for nitriding of the martensitic stainless steel AISI420-J2 specimens at three different temperatures (653 K, 673 K, and 693 K) with use of the hollow cathode setup. The influence of nitrogen supersaturation on the lattice expansion, the phase transformation and the grain refinement are investigated. In particular, the effect of lattice expansion and high plastic straining on the formed phases, the phase transformation mechanisms and the grain size refinement are precisely analyzed. High density, low temperature plasma nitriding provides a new means to improve the microstructure and mechanical properties of martensitic stainless steel molds.

2. Experimental Procedures

A high-density plasma nitriding system consisted of the vacuum chamber, the evacuation system, the DC-RF generators working in the frequency of 2 MHz, the gas supply of N2 and H2, and the heating unit located under the cathode plate as illustrated in Fig. 1. The thermocouple was inserted into this cathode plate to monitor the holding temperature. In the vacuum chamber, the specimen was placed inside a hollow cathode setup on the cathode plate, which was electrically connected with DC generator. After the previous study,24) the plasma density was enhanced in the inside of setup to attain high nitrogen ion density up to 3×1017 ions/m3.25) This hollow cathode setup includes a rectangle-shaped tube with the size of 40×20×70 mm3 and the thickness of 2 mm. After quantitative plasma diagnosis on the ion density distribution,26) the specimen was placed at the position with the highest nitrogen ion density, far half from the gases inlet, as shown in Fig. 1. All process parameters are summarized in Table 1. Detailed nitriding procedures were explained elsewhere.24)

Fig. 1.

Illustration of the high-density plasma nitriding system.

Table 1. Standard processing parameters for the present plasma nitriding.
ParameterPre-SputteringPlasma nitriding
Temperature (K)693, 673, 653693, 673, 653
Pressure (Pa)30060
RF-Voltage (V)250250
DC-Voltage (V)350500
Gas Flow (ml/min)160 N2160 N2, 30 H2
Gas ratio (%)100%N284%N2+16%H2
Duration (sec)180014400

The martensitic stainless steel AISI-SUS 420-J2 was employed as a test-piece; its chemical composition was C = 0.15%, Mn = 1%, P = 0.04%, S = 0.03%, Ni = 0.6%, Si = 1%, Cr = 14%, and Fe in balance. A circular disk specimen with a diameter of 25 mm and thickness of 5 mm was prepared.

All nitrided specimens were analyzed by the X-ray diffractometer (XRD). XRD was performed by Rigaku SmartLab with the monochromatic CuKα radiation (λ = 0.1542 nm) and Bragg–Brentano geometry (40 kV, 30 mA). In measurement, 2θ ranged from 30° to 90° with the scanning speed of 5 mm/min and the step angle of 0.02°. The microstructure of the nitrided layer was analyzed by scanning electron microscope (SEM).

After preliminary polishing, scanning electron microscope (SEM; HITACHI SU-70) was used to obtain the high-resolution images of the microstructure for the specimen. First, the specimen is attached to the specimen holder device and inserted into a high vacuumed chamber after ensuring that the total height is 25.4 mm ≈1 inch. Then the electron beam was started at the accelerating voltage of 5 KV with a working distance from the analyzed surface of 15 mm. A YAGBSE device was used for controlling the contrast as well as the focus and adjusting the electron beam alignment to the area to be analyzed. Then energy dispersive spectroscopy (EDS) device and software were utilized to obtain distribution mapping over a specified depth for nitrogen, chromium, iron, and carbon. Finally, the electron backscattering diffraction (EBSD) was utilized with the accelerating voltage of 20 kV, the working distance of 20 mm, the magnification (2000), and the resolution of 0.1 μm. After this EBSD, the inverse pole figure (IPF) was determined for each constituent grain on the cross-section of test-piece. In addition, the kernel average misorientation (KAM) and the phase mapping, were also measured respectively to describe the straining and phase transformation processes.

Microhardness was measured on the surfaces and cross-sections for all the specimens. The microhardness testing machine (Mitsutoyo HM-200) was used by applying the load of 10 g or 0.1 N for surface measurements and the load of 50 g or 0.5 N for cross-sections with 10 μm steps, respectively.

3. Results

Figure 2(a) shows the SEM micrograph of the nitrided specimen at 673 K for 14.4 ks. The microstructure becomes different across the line A. The original grains are seen below the line A; these grains turned to be fine above this line A. The nitrogen mapping by EDS is also shown in Fig. 2(b). High nitrogen content is detected above this line; the measured content abruptly decreases across this border. This surface layer in Fig. 2 is identified as a nitrided layer with the thickness of 85 μm. After Kim et al.,21) the deep nitrided layer with the thickness of 20 μm was reported in the plasma nitriding of AISI420 substrates at 673 K for 14.4 ks. Even through the low temperature nitriding process, the significantly thick nitrided layer is formed for shorter duration time than those for high temperature (>773 K) nitriding processes in 7, 9, 10). In addition, the grain size is significantly refined in this nitrided layer by the present plasma nitriding while no reports were found on this microstructure refinement even in 21).

Fig. 2.

(a) SEM cross-sectional micrograph of the plasma nitrided AISI-420J2 specimens for 14.4 ks at 673 K (b) EDS nitrogen mapping on the cross-section of the plasma nitrided AISI420-J2 specimen at 673 K for 14.4 ks.

Figure 3 represents the XRD diagrams both for the as-received (bare) and the plasma nitrided specimens for 14.4 ks at 653, 673, and 693 K, respectively. The iron nitrides of ε-Fe3N and γ′-Fe4N phases were present on the surface but CrN was present in a trace level. These nitrides are formed only at the vicinity of the specimen surface; no nitride peaks are detected in XRD diagrams at the depth of 20 μm from the surface.

Fig. 3.

XRD diagrams of martensitic stainless steel AISI 420-J2 specimens.

The original martensitic stainless steel is characterized by the high-intensity peaks for α′(110), α′(200), and α′(211) at 2θ = 44.5°, 64.84°, and 82.14°, respectively. The first and second peaks are shifted to a lower angle of 2θ = 43.8°, 63.78° respectively after nitriding. This peak-shift of α-phase implies that the original α-phase martensitic lattice expands itself by super-saturation with higher nitrogen content than the maximum solubility limit of nitrogen solute of 0.1 mass%. That is, the expanded martensite α′N is formed by the low temperature plasma nitriding as suggested by Kim et al.21) Expanded austenite for γN (111) and γN (200) are also detected at 2θ = 41.2°, 48° respectively.

The average lattice strain by this lattice expansion is calculated from the measured lattice constants before and after nitriding to be 1.6% as listed in Table 2. The nitrogen concentrations for all specimens are measured by EDS with a precision of ± 2 at%.

Table 2. Summary of the nitriding treatment conditions and results.
Nitriding specimen’s conditionsNitrided Thickness (μm)Nitrogen surface Concentration (at%)Surface Hardness (HV0.1)Lattice Parameter (nm)Lattice Expansion (%)Phases
As received specimen2200.28747α
Specimen nitrided at
(653 K, 14.4 ks)
782715300.291821.51αN, γN,
ε, γ′ (at surface)
Specimen nitrided at
(673 K, 14.4 ks)
853117400.292011.58α′N, γN,
ε, γ′ (at surface)
Specimen nitrided at
(693 K, 14.4 ks)
762915900.29231.68αN, γN,
ε, γ′ (at surface)

The nitrogen content distribution in depth as well as the hardness depth profile are shown in Fig. 4 for three nitrided specimens at 693 K, 673 K and 653 K. In all three cases, high hardness is attained at the vicinity of surface, slightly decreases in the depth down to the matrix hardness toward the nitriding front end; e.g., at 653 K, the hardness became 1530 HV near the surface, slightly decreases down to 220 HV toward the nitriding front end of 78 μm.

Fig. 4.

a) Microhardness depth profile b) Nitrogen content depth profile of the cross-section of the plasma nitrided AISI420-J2 specimens at three different temperatures of 653 K, 673 K, and 693 K. Nitrogen concentrations are measured by EDS.

Irrespective of the holding temperature, high surface nitrogen contents more than 25 at% are present near the surface, decreases down to 10 at% at 20 μm depth as shown in Fig. 4(b). In the middle of depth, a plateau with constant nitrogen concentration by 8–10 at% was observed toward the nitriding front end.

The inverse pole figure (IPF), the kernel average misorientation (KAM), and the phase mapping were analyzed by EBSD and shown in Figs. 5(a), 5(b), and 5(c) respectively, for the nitrided specimen at 653 K for 14.4 ks. From the measured crystalline orientation, each constituent grain is identified as a color mapping in Fig. 5(a). From this mapping, the microstructure in the depth of nitrided AISI420J2 is classified into three regions.

Fig. 5.

Electron backscattered diffraction (EBSD) crystallographic images of martensitic stainless steel AISI420-J2 specimen nitrided at 653 K for 14.4 ks. (a) the IPF map of the grain size, (b) the KAM strain mapping, and, (c) the mapping of the austenitic and martensitic phases.

At the vicinity of surface down to 20 μm or in the region-I, the original coarse grains are refined to have the average grain size of 0.15 μm in nearly equal to the spatial resolution by the present EBSD analysis. In the middle range of depth from 20 to 70 μm below the line-A, the original grains are partially refined to the average grain size of 2.5 μm. Beyond the nitriding front end or the line-B in Fig. 5(b), or, in the region-III, the grain size remains to be the same as the original size of 10 μm even after plasma nitriding.

This crystalline refinement in the nitrided layer thickness is unique to this low temperature plasma nitrided stainless steel matrix, never reported in the previous studies.7,11,21) All average grain sizes for all specimens are calculated according to ASTM standard as listed in Table 3.27)

Table 3. Average grain size and austenitic phase transformation for AISI420-J2 specimens in three different regions.
Specimen IdentificationsAverage Grain Size Diameter (μm)Austenitic phase transformation (%)
Region IRegion IIRegion IIIRegion IRegion IIRegion III
As received specimen9.4100% martensite
Specimen nitrided at (653 K, 14.4 ks)0.152.510.4344.60.1
Specimen nitrided at (673 K, 14.4 ks)0.24.110.233.92.10.1
Specimen nitrided at (693 K, 14.4 ks)0.15210.828.23.80.1

In Fig. 5(b), KAM represents the intergranular misorientations among the neighboring grains. When this KAM is nearly zero, all grains have no strains in their inside and no distortion was seen on each grain boundary.

In the region-I, high KAM distributed just in this region together with the phase transformation from martensite to austenite with high volume fraction of 34% as seen in Fig. 5(c). This high misorientation angle data reduced but distributed as a network in region-II. The phase transformation took place only at the high KAM networks with the small volume fraction of 4.6% as listed in Table 3. In the region-III in Figs. 5(b), 5(c), KAM was less than 1 degree; the original grains were not strained and the grain boundaries remain the same as the original coarse grains. Moreover, there is no austenitic phase transformation in this region.

4. Discussion

After survey in the literature, the plasma nitriding processes of martensitic stainless steels as well as their results are compared in Table 4.5,7,9,11,21,28,33,34) In case of the nitrogen ion implantation, the nitrided layer thickness was only 5 μm with the anisotropic lattice expansion up to 3.5%. High energetic implantation of nitrogen ions at 25 kV induced this high lattice strain. In case of those RF, DC and glow discharged plasma related processes, formation of thicker nitrided layer from 10 to 90 μm was reported with strong dependency on the nitriding time and holding temperature. Although a little lattice expansion was reported in them, both wear and corrosion resistance of the material was improved as well as the hardness increase via those processes. Increase of the nitriding temperature and time not only reduces the corrosion resistance of stainless steels but also raises their industrial application costs. In the present study with high density plasma nitriding using the hollow cathode setup, the nitrided layer thickness reaches 85 μm even after plasma nitriding at 673 K for 14.4 ks. High density nitrogen atom flux from the plasmas is responsible for nitrogen super-saturation into the martensitic matrix with high content even at lower temperature.

Table 4. Comparison of nitriding conditions and results among different research studies.
ReferencesTreatment systemsStainless steelsTemperature (K)Time (h)Nitrogen Surface Content (at%)Nitrided Thickness (μm)
[34]Nitrogen ion implantationAISI420653 K1 h10 at%5 μm
[7]DC plasma nitridingAISI416673 K4 h14 μm
[21]RF plasma nitridingAISI420673 K4 h20 μm
[9]plasma nitridingCA-6NM773 K2 h25 μm
[33]glow discharge plasma nitridingAISI420803 K20 h17 at%65 μm
[5], [11]DC-pulsed glow dischargeAISI420623 K15 h90 μm
[28]Ion nitridingAISI410673 K120 h20 at%160 μm

No nitrides were formed during this low temperature plasma nitriding process; most of nitrogen atoms are unbound with constituent iron and chromium atoms. After the first principle calculation in 29, 30), the nitrogen super-saturation accompanies with the occupation process of octahedral sites in the lattice with the interstitial nitrogen atoms. In a similar discussion on the interaction between two nitrogen interstitial atoms and one vacancy,29) two possible configurations are considered: occupation of vacancies by nitrogen interstitial atoms at the absence of body center vacancy (BCV) or in the presence of BCV.

In the former, two interstitial nitrogen atoms stay into Fe/Cr superlattice edges; two octahedral lattice sites are occupied with the largest distance and the least N–N repulsive energy, as shown in Fig. 6(a). The calculated nitrogen content in this configuration is estimated to be 11 at%, which is consistent with the measured nitrogen content regarding the plateau of nitrogen concentration in Fig. 4(b). In the latter, the nitrogen occupation process takes place at two first neighboring octahedral sites with the highest binding energy aligned along <100> directions. As shown in Fig. 6(b), the Fe/Cr bcc structure with BCV has two nitrogen atoms in its inside; the estimated nitrogen content reaches to 50 at%.

Fig. 6.

Two possible configurations for occupation process of nitrogen atoms into Fe/Cr matrix. (a) Iron superlattice with the absent of a vacancy, (b) Single iron lattice with the presence of body center vacancy (BCV).

This demonstrates that nitrogen super-saturation process into Fe/Cr matrix is driven by the occupation process of the octahedral lattice vacancy sites by the large amount of nitrogen interstitials. This attraction between vacancies and nitrogen atoms also influences on the nitrogen diffusion process.

In the super-saturated matrix, the nitrogen atom works as an alloying element to stabilize the austenitic structure by lowering the Ms point. After Imai, et al.,31) high nitrogen concentration in the austenitic iron powders lowered the Ms point as well as the free energy change curve from austenite to martensite. Blawert et al.32) also reported that the ferrite phase of duplex austenitic-ferritic steel transformed to the expanded austenite by the plasma ion implantation and immersion (PIII) process. These two findings suggest that high nitrogen supersaturation increases the free energy by elastic lattice straining to initiate the phase transformation from the expanded martensite to the expanded austenite. That is, the expanded martensite with high strain energy in its lattices is ready to transform into the austenite during the high nitrogen super-saturation process.

As suggested by 35), the KAM data represent the equivalent plastic strain; e.g., highly angled KAM data up to 5 degrees represent high distortion of grains and grain boundaries. This high distortion generates the true plastic straining and spin-rotation of grains. The original coarse grains of AISI420-J2 stainless steels are subjected to high straining and rotation and refined into smaller grains.

As suggested by 15, 25), the high temperature plasma nitriding was governed by the nitrogen diffusion process as well as the precipitation reaction at the nitriding front end. The nitrogen content decreases exponentially from the maximum nitrogen content of 0.1 to 0.2 mass%. In case of the low temperature nitriding, the nitrogen content depth profile is nearly constant except for the initial decay from the surface to the plateau and for the final decrease toward the nitriding front end. Many researchers tried to describe this difference between the high and low temperature plasma nitriding behavior.

After the previous studies,36,39) the nitrogen diffused much faster in the expanded phases than in the original crystal structure. This difference in the diffusion rate resulted in the nitrogen depth profile different from the conventional error function, where the nitrogen diffusion only governs the nitrogen super-saturation process. Those studies stood on the interaction between the constituent elements and the nitrogen interstitials to explain the above nitrogen diffusion rate. Some researchers believed that the chromium (Cr) lattice sites have a role in trapping/detrapping of nitrogen atoms. Once all chromium trap sites are supersaturated with one or more nitrogen atoms, any additional nitrogen atoms can diffuse deeper until the end of the expanded supersaturated surface layers.36,37,38) Mändl and Rauschenbach prove the impracticality of the Cr trapping diffusion model.39) Instead, they demonstrated that the non-error function diffusion mechanism is directly related to the nitrogen concentration in each region. Higher N-diffusivity induced at higher nitrogen surface concentrations are corresponding to larger lattice constants in the expanded phases. A change in the electron density distribution with increasing lattice constant could provide a quite rapid increase in N-diffusivity as the potential barrier is lowered during this process.

The present experiment reveals that both models cannot explain the nitrogen content plateau in the depth profile. The occupation process of octahedral lattice vacancy sites in the bcc-structured Fe/Cr lattice with the nitrogen interstitial atoms must be taken into account even in the diffusion process.

5. Conclusion

Through high density plasma nitriding on martensitic stainless steel AISI420-J2 at low-temperatures of 653, 673, 693 K, significant mechanical and metallurgical advantages are obtained in what follows:

(1) Formation of expanded austenite (γN) and expanded martensite (α′N) phases with lattice strain up to 1.6% in average by the nitrogen supersaturation,

(2) High nitrogen concentration up to 31 at% in the surface region, partially including the bound nitrogen in the iron nitrides,

(3) Nearly constant nitrogen content in depth by 10 at% except for the initial decrease from the surface and for the final decay toward the nitriding front end,

(4) High hardness up to 1500–1800 HV at the surface by this nitrogen supersaturation,

(5) Grain size refinement from the original size of 10 μm down to 0.2 to 0.1 μm by the high plastic distortion as the slip-line field, following the elastic lattice straining,

(6) Phase transformation from martensite to austenite by the volume fraction of 34% in average,

(7) Possibility of nitrogen diffusion process advancing in parallel with the occupation process of octahedral sites in the bcc-structured Fe/Cr supercell by the nitrogen atoms.

Acknowledgements

The authors would like to express their gratitude to Mr. T. Yoshino (Komatsu-Seiki Kosakusho, Co. Ltd.) and Mr. Hiroshi Morita (Nano Film & Coat Laboratory, LLC) for their help in measurement. This study is financially supported by the ABE-Initiative Program, Japanese Government and the METI-program on the supporting industries in 2017.

References
 
© 2018 by The Iron and Steel Institute of Japan
feedback
Top