Microstructure and Fracture Toughness of Nitrided D2 Steels Using Potential-Controlled Nitriding

: Potential-controlled nitriding is an effective technique for enhancing the life of steel molds and dies by improving their surface hardness and toughness against fatigue damage. In this study, the effect of the nitriding potential on the microstructure and fracture toughness of nitrided AISI D2 steels was investigated. The nitrided layers were characterized by microhardness measurements, optical microscopy, and scanning electron microscopy, and their phases were identiﬁed by X-ray and electron backscatter diffraction. As the nitriding potential increased to 2.0 atm − 1/2 , an increase in the surface hardness and fracture toughness was observed with the growth of the compound layer. However, both the surface hardness and the fracture toughness decreased at the higher nitriding potential of 5.0 atm − 1/2 owing to the increased porosity in the compound layers, which mainly consist of the ε (Fe 2–3 N) phase. Additionally, by observing crack growth behavior, the fracture toughness was analyzed considering the material characteristics of the diffusion and compound layers. The fracture toughness was inﬂuenced by the location of the initial Palmqvist cracks due to the localized plastic deformation of the diffusion layer and increased crack length due to the porous compound layer. This study showed that potential-controlled nitriding can be extended to molds and dies and can improve their mechanical properties, prolonging their service lives.


Introduction
High-carbon, high-chromium AISI D2 steel, also known as cold work tool steel, is widely used in the fabrication of punches, shear knives, tools, molds, and dies because of its hardenability and high resistance to wear and softening [1,2]. Recently, it has been utilized for various manufacturing applications to enhance the working life of components against impact, shear, and fatigue damage [3][4][5]. For instance, cold work tool steel performance was improved by modifying its chemical composition to obtain high strength, toughness, and high fatigue resistance [3]. Heat treatment processes, such as cryogenic treatment and salt bath quenching, were also developed to further improve the hardness, toughness, and wear resistance of cold work tool steel [4,5].
Well-known techniques for surface modification, such as nitriding treatment (e.g., compound-free, ion-nitriding) and physical vapor deposition (e.g., TiN, AlTiN coatings), have been studied and implemented to improve the working performance of molds and dies [6][7][8]. Nitriding is one of the most useful thermochemical surface treatments because it can be performed at temperatures of 450-600 • C, resulting in low distortion and high surface hardness [9,10]. However, there are deleterious effects on the mechanical properties of the nitrided layers, namely decreased surface hardness, wear resistance, and toughness [8,11]. This is due to the presence of excess nitrogen and discontinuous precipitation, which result in the coarsening of the nitrides and porosity of the compound layer during nitriding [12]. Techniques such as ion-nitriding and gas-controlled nitriding have been developed to guarantee the long service life of the nitrided layer [13][14][15].
Mechanical properties, such as high surface hardness and wear resistance, combined with toughness, are required to enhance the durability of steel dies and molds [16,17]. As reported by Yong et al. [18], the low toughness of the compound layer leads to brittle cracks and decreases the die and mold thickness owing to wear damage, resulting in the cracking and spalling of the compound layer. Nolan et al. [19] applied plasma nitrided treatment on AISI H11 steel to evaluate the fracture toughness of the compound layers as a function of indentation load using the Vickers hardness test and established that fracture toughness is dependent on the compound layer thickness owing to different crack behaviors. The elastic and toughness properties of the nitrided layers were recently investigated using micro-mechanical tensile test methods [20]. To improve the life of die and mold steel, a controlled nitriding process that involves regulating the applied nitriding potential K n is crucial for minimizing the crack growth of the nitrided layer [21].
As mentioned earlier, the fracture toughness of nitrided AISI D2 steel relies heavily on the material characteristics of the nitrided layers. In other words, the thickness of the compound layer, as well as the phase evolution of the nitrided layer, must be evaluated in order to accurately estimate the fracture behavior, which can be achieved by controlling the nitriding potential [22]. However, studies on this topic using die and mold steels are scarce. Herein, we investigated the effect of the nitriding potential on the microstructural evolution and fracture toughness of nitrided AISI D2 steel. This steel was chosen as it is widely applied in the gas nitriding process to prolong die and mold working life. The crack growth behavior of the nitrided layers is analyzed in detail to verify its association with fracture toughness.

Materials and Methods
The chemical composition of the AISI D2 steel used is listed in Table 1. Specimens of size Ø30 mm × 10 mm were prepared and subjected to the following heat treatment procedure to obtain a hardness of approximately 56 HRC. The specimens were austenitized at 1035 • C in a vacuum furnace, quenched in N 2 gas at 3 bar pressure, and double-tempered for 90 min at different temperatures of 550 and 560 • C. Prior to the gaseous nitriding process, the quenched and tempered (QT) AISI D2 specimens were ground to 1500 grits using several grades of SiC sandpaper and cleaned with ethanol in an ultrasonicator bath. Gas nitriding was then performed in a nitriding furnace (SECO/WARWICK, HRNe606090), as illustrated schematically in Figure 1. Pre-heating was conducted at 350 • C; this was followed by gas nitriding in NH 3 gas at 520 • C for 10 h. The gas atmosphere during nitriding was controlled using a hydrogen sensor.  Then, the nitriding potential K n , which defines the chemical potential of nitrogen related to the evolution of the nitrided layer, was calculated using the H 2 gas fraction. To investigate the effect of nitriding potential, the nitriding potentials were kept stable at 0.2, 1.0, 2.0, and 5.0 atm −1/2 for each nitriding process as per AMS 2759-10B. The nitriding process can be expressed by the decomposition reactions of nitrogen from NH 3 gas as follows: where P NH 3 is the partial pressure of ammonia, and P H 2 is that of hydrogen. After the nitriding treatment, the surface and section hardnesses were obtained using a Vickers micro-hardness tester (Future Tech, FLV-10 ARS-F, Kawasaki, Japan) with a load of 100 gf for a dwell time of 10 s. At least five measurements at each section were taken, and the average hardness value was obtained. The effective hardening depth was determined as the depth corresponding to the high 50 HV 0.1 from the hardness of the quenched and tempered specimen as per ISO 18203.
To characterize the microstructure, the nitrided specimens were etched with a mixture of 100 mL ethanol and 3% nitric acid solution. The cross-sectional microstructures were then observed using optical microscopy and scanning electron microscopy (SEM, FEI, NNS450, Hillsboro, OR, USA). To determine the phase evolution of the nitrided specimens, X-ray diffraction (XRD, PANalytical, X'Pert-PRO MPD, Malvern, UK) was performed under accelerating conditions of 40 kV and 30 mA and diffraction angle measurements from 20 • to 90 • using the Cu-Kα target were taken. Electron backscatter diffraction (EBSD, Hitachi, SU5000, Tokyo, Japan) was performed to characterize the phase of the compound layer. Finally, hardness measurements were carried out using a Vickers macrohardness tester (Mitutoyo HV-100, 810-440 K, Kawasaki, Japan) with a load of 50 kgf for a dwell time of 5 s, and the Vickers hardness (HV) was calculated using Equation (3) [23]: where P is the indentation load (kgf), and d m is the mean indentation diagonal (mm). To determine the fracture toughness of the nitrided layer, the Palmqvist-method was used by estimating the total crack length. The Palmqvist crack lengths at the indentation corners were measured using SEM at 500× magnification. The fracture toughness was calculated using Equation (4) [23]: where T is the total crack length (mm). Figure 2 shows the hardness profiles of the QT material and nitrided specimens in relation to the distance from the surface. The average hardness of the heat-treated specimen was 570 HV 0.1 . With a nitriding potential of K n = 0.2 atm −1/2 , the hardness was approximately 1130 HV 0.1 on the surface of the nitrided layer, which decreased as the depth increased. The case depth at K n = 0.2 atm −1/2 was measured to be approximately 100 µm from the top surface. With a nitriding potential of K n = 1.0 atm −1/2 , the surface hardness and the case depth increased to about 1260 HV 0.1 and 130 µm, respectively. As K n increased to 2.0 atm −1/2 , the surface hardness reached a maximum value of approximately 1270 HV 0.1 , and there was no significant change in the case depth with the increase in nitriding potential. However, at K n = 5.0 atm −1/2 , the decrease in surface hardness was observed to be approximately 960 HV 0.1 . The compound layer grew with the increase of the nitriding potential to 2.0 atm −1/2 . However, the hardness profile at K n = 5.0 atm −1/2 deviated from the general shape.

Microstructure
The cross-sectional microstructures of the QT and nitrided AISI D2 specimens are shown in Figure 3. In the QT base material, primary carbides were observed in tempered martensite owing to quenching and tempering. With a nitriding potential of K n = 0.2 atm −1/2 , the compound layer was finely formed till about 0.7 µm from the top surface ( Figure 3a). In particular, at the nitriding potential of K n = 1.0 atm −1/2 , the compound layer grew to about 2 µm, and chromium nitrides (CrN) were distinctly observed in the diffusion layer ( Figure 3b). With increasing nitriding potentials (K n = 2.0 atm −1/2 to K n = 5.0 atm −1/2 ), the compound layer remarkably grew from about 7.8 µm to 8.7 µm (Figure 3c,d, respectively).    Figure 5 shows the EBSD of the nitrided AISI D2 specimens for characterizing the phase of the nitrided layers. Prior to phase analysis, the phases of the compound layer can be easily distinguished to consist of ε (Fe 2-3 N) and γ' (Fe 4 N) phases, which have hexagonal close-packed (HCP) and face-centered cubic (FCC) structures, respectively [24]. At the lower nitriding potential (0.2 atm −1/2 ), the α-Fe phase was detected on the surface of the nitrided layer, and there were no obvious nitride phases consisting of ε (Fe 2-3 N) and γ' (Fe 4 N) near the nitrided surface. At K n = 1.0 atm −1/2 , the ε (Fe 2-3 N) phase, which consisted of the compound layer, was mainly observed, and the γ' (Fe 4 N) phase was finely detected around the compound layer. As the nitriding potential increased from 2.0 to 5.0 atm −1/2 , the compound layer, which mainly consisted of the ε (Fe 2-3 N) phase, grew remarkably on the surface of the nitrided AISI D2 specimens. From both XRD and EBSD phase analyses, it was found that the compound layer of the nitrided AISI D2 specimens mainly consisted of the ε (Fe 2-3 N) phase.  Figure 6 shows the fracture toughness (K Ic ) as a function of various nitriding potentials compared with the surface hardness and compound layer thickness. With a nitriding potential of K n = 0.2 atm −1/2 and a compound layer of 0.7 µm thickness, the fracture toughness and the surface hardness were observed to be about 8.5 MN/m 3/2 and 1130 HV 0.1 , respectively. As the nitriding potential increased to 2.0 atm −1/2 , the compound layer thickness increased, and the maximum fracture toughness and surface hardness were measured to be approximately 12.6 MN/m 3/2 and 1270 HV 0.1 , respectively. However, at the higher nitriding potential of K n = 5.0 atm −1/2 , the fracture toughness decreased to approximately 10.0 MN/m 3/2 with the decrease in the surface hardness despite the compound layer thickness increasing to 8.7 µm. Thus, the fracture toughness was found to be related to surface hardness.

Crack Characteristics
To examine the effect of the nitriding potential on the fracture toughness with the evolution of the nitrided layer, the crack characteristics were investigated through Vickers indentation. Figure 7 shows SEM images of the Palmqvist cracks and ring-shaped cracks. With the increase in nitriding potential (K n = 0.2 to K n = 2.0 atm −1/2 ), the Palmqvist crack length gradually decreased to 45 µm with the growth of the compound layer. However, at K n = 5.0 atm −1/2 , the crack length increased again to approximately 70.4 µm, despite the growth of the compound layer to approximately 8.7 µm. In addition, around the indentation edges, the length of the ring-shaped cracks decreased with the growth of the compound layer as K n increased from 0.2 to 2.0 atm −1/2 , which became zero (no ring-shaped cracks) at K n = 5.0 atm −1/2 .  Figure 8 shows the crack patterns occurring in the compound-free nitrided layer, obtained by polishing the compound layer on the nitrided surface (K n = 0.2 atm −1/2 ). As observed in Figure 8a, the ring-shaped cracks occurred around the indented area due to the plastic deformation in the compound-free nitrided layer, and semi-circular cracks were formed beneath the indentation. As shown in Figure 8b, micro-cracks were formed along the primary carbides when the Palmqvist crack propagated outward along the diagonal of the indentation.

Discussion
In this study, we investigated the effect of the nitriding potential on the evolution of the nitrided layers, and consequently, the fracture toughness by analyzing the crack propagation. The results indicate that the nitriding potential has a significant effect on fracture toughness. In addition, a previous study [19] showed that the fracture toughness is dependent on the thickness of the compound layer; however, in this study, it was verified that phases, such as the chromium nitrides (CrN) and iron nitrides (ε-Fe 2-3 N), and the porous compound layer significantly influenced the different behaviors of crack growth that resulted from the indentation. It was found that these were the main factors affecting fracture toughness and surface hardness.
As shown by the SEM images in Figure 9a-f, with increasing nitriding potential (K n = 0.2-5.0 atm −1/2 ), the initial Palmqvist crack moves nearer to the indentation corners with the growth of compound layers. At the lower nitriding potential (0.2 atm −1/2 ), the semi-circular cracks were formed beneath the Vickers indentation (see Figure 9a). Subhash et al. [25] reported that micro-cracks are developed by localized plastic deformation (shear bands) owing to the Vickers indentation, which then grows into semi-circular cracks. The crack characteristics of nitrided H11 steel are dependent on the thickness of the compound layer, as reported by Nolan et al. [19].
For the crack characteristics of the nitrided mold steel, alloy compositions (Cr, Mo, V, etc.) affect the precipitation reactions of the diffusion layer. During the nitriding process, M 7 C 3 gradually decomposes with decarburization as CrN grows. In this recrystallization, the M 7 C 3 retains discontinuous morphologies of carbide [26][27][28]. In particular, as shown in Figure 10, a discontinuous precipitation network of newly formed carbides (M x C y ) is present in the carbon-rich zone along the coarse CrN nitrides. As the localized plastic deformation occurred during the Vickers indentation, the interface brittle cracks, as shown in Figure 8, developed at the weakly bonded interfaces between the different precipitates. This crack formation during the localized plastic deformation of the diffusion layer was found to be influenced by the initial Palmqvist cracks. Accordingly, at the lower nitriding potential (K n = 0.2 atm −1/2 ), the fracture toughness was the lowest at 8.5 MN/m 3/2 when the crack length was 80.2 µm.  As shown in Figure 9a-f, with increasing nitriding potential (K n = 0.2-5.0 atm −1/2 ), the initial Palmqvist cracks move closer to the indentation corners with the growth of compound layers. This eventually leads to a decrease in the length of the Palmqvist crack to approximately 45.5 µm owing to the differently located initial cracks (see Figures 7c and 9c). Therefore, at the nitriding potential of 2.0 atm −1/2 , the fracture toughness was found to be the highest at approximately 12.6 MN/m 3/2 . However, at the higher nitriding potential of 5.0 atm −1/2 , the crack length increased again to approximately 70.4 µm as the compound layer grew to approximately 8.7 µm (Figure 7d). This may be related to the porous layer formed near the surface of the compound layer (Figure 9f) due to the weakly bonded dissolved nitrogen at the higher nitriding potential [29,30]. These porosities near the surface of the compound layer led to a decrease in the surface hardness owing to stress concentrations during the Vickers indentation. Thus, the increase in the Palmqvist crack length was influenced by the porosity of the compound layer, and thereby, the consequent fracture toughness decreased to approximately 10.0 MN/m 3/2 .
In the diffusion layers developed by the nitriding potential, the discontinuous precipitation network of the carbides (M x C y ) grew in the carbon-rich zone along the coarse CrN nitrides. Furthermore, at a higher nitriding potential, the compound layer consisting of the ε (Fe 2-3 N) phase grew and developed pores near the surface of the compound layer due to excess nitrogen. These material characteristics are closely related to surface hardness and fracture toughness. Thus, the crack growth behavior can be briefly described by a schematic diagram (Figure 9g,h) of the crack patterns in the nitrided AISI D2 specimens. This figure shows the location of initial cracks according to the localized plastic deformation in the diffusion layer and the increase in the crack length due to the porous compound layer. With the low nitriding potential of K n = 0.2 atm −1/2 , the ring-shaped cracks grew radially out toward the indentation due to the plastic deformation of the diffusion layers and inside the indentation; micro-cracks grew and developed into Palmqvist cracks due to the interface of the weakly bonded precipitates. Therefore, the fracture toughness was the lowest at the nitriding potential of 0.2 atm −1/2 . However, at K n = 5.0 atm −1/2 , the porosities developed on the surface of the compound layer led to an increase in the crack length ( Figure 9h). Thus, the control nitriding for D2 steel can be effectively used to form the non-porous compound layer, which leads to the increase in fracture toughness by controlling the K n .

Conclusions
In this study, the evolution of nitrided layers and fracture toughness of AISI D2 steels were investigated using various nitriding potentials.

•
The average hardness of the heat-treated specimen was measured to be 570 HV 0.1 , whereas that of the nitrided steel peaked at 1270 HV 0.1 for applied nitriding potentials below 5.0 atm −1/2 . At the high nitriding potential of K n = 5.0 atm −1/2 , the surface hardness decreased to about 960 HV 0.1 . • Chromium nitrides (CrN) develop in the diffusion layer as the carbide precipitation network grows in the carbon-rich zone owing to the recrystallization resulting from nitrogen diffusion. At a nitriding potential of 5.0 atm −1/2 , a compound layer consisting of the ε (Fe 2-3 N) phase grows and develops pores near the surface of the compound layers due to the excess nitrogen. • The fracture toughness was measured to be approximately 8.5 MN/m 3/2 at K n = 0.2 atm −1/2 at which the compound layer had a thickness of about 0.7 µm. As K n was increased to 2.0 atm −1/2 , the compound layer further increased in thickness, and the fracture toughness peaked at 12.6 MN/m 3/2 . However, as K n was further increased to 5.0 atm −1/2 , the fracture toughness decreased to approximately 10.0 MN/m 3/2 with a decrease in the surface hardness, despite the compound layer growing up to 8.7 µm thick. Thus, the fracture toughness was found to be related to the surface hardness. • Two main reasons for different crack growth behaviors of the nitrided layers with changes in the nitriding potentials were observed: the location of the initial crack inside the indentation due to the localized deformation of the diffusion layer and an increased crack length due to the porous compound layer. These main factors affected the fracture toughness. In short, with the growth of the non-porous compound layer, the increase in fracture toughness was closely related to the location of the initial crack.
This study showed that potential-controlled nitriding can be extended to molds and dies and can improve their mechanical properties, prolonging their service lives. Data Availability Statement: The data are available from the authors upon reasonable request.