Figure shows the EPMA image and the corresponding elemental maps of the Ti-4Al-6V substrate. The α-Ti grains (dark area in Figure ) were rich in Al, and the intergranular β-Ti (white islands in Figure ) was rich in V. This is due to the fact that Al is an α stabilizer, while V is a β stabilizer.
Figure shows the XRD patterns of the Ti-4Al-6V alloys nitrided at 850°C. After nitriding for 1 h, a distinct, tetragonal Ti
2N layer was formed on the α-Ti-rich matrix (Figure ). After nitriding for 6 h, the α-Ti-rich matrix peaks became weaker and weak fcc-TiN peaks appeared, owing to the increased nitriding time (Figure ). Here, Ti
2N began to exhibit a preferred orientation along the (002) direction. After nitriding for 12 h, the α-Ti matrix peaks disappeared, whereas weak TiN peaks and strong Ti
2N peaks with a (002) preferred orientation appeared (Figure ). Hence, it is seen that the α-Ti-rich matrix transformed into Ti
2N from the surface and later, into TiN as the nitriding progressed. Ti
2N exists in a narrow range of approximately 34 at.%N, while TiN displays a wide range of nitrogen solubility above 38 at.%N at 850°C in the Ti-N phase diagram. The formation of Ti
2N indicates that the minimum nitrogen content of 34 at.%N is attained after nitriding for 1 h (Figure ). The Ti-6Al-4V alloy exhibits an allotropic transition between the low-temperature hcp α-Ti and the high-temperature bcc β-Ti at 995°C. Since nitriding was performed at a temperature lower than the β-transus temperature, α-Ti, TiN, and Ti
2N were detected in Figure . When the Ti alloys were nitrided at 950°C to 1, 050°C for 1 to 5 h in atmospheric nitrogen, surface layers of TiO
2, TiN, Ti
2N, and α-Ti with N in a solid solution (viz. α-Ti(N)) were formed [
1-
3]. In this study, the residual oxygen was well regulated so as not to form TiO
2.
Figure shows the SEM images of the nitrided Ti-4Al-6V alloys. Regardless of the nitriding time, all of the surfaces exhibited a golden yellow color and consisted of fine nitrides with a smooth surface. The nitrided Ti-4Al-6V alloys in all of the cross-sectional images consisted of an outer Ti-N compound layer, inner α-Ti(N) diffusion zone, and matrix. According to Figure , the compound layer consisted of Ti
2N with and without TiN, and the diffusion zone consisted of α-Ti having dissolved nitrogen. It is noted that α-Ti can dissolve up to 7.6 wt.% nitrogen. After nitriding for 1 h, the thicknesses of the Ti-N layer and Ti(N) zone were 0.8 and 2.5 μm, respectively (Figure ). After nitriding for 6 h, their thicknesses were 2.3 and 4.6 μm, respectively (Figure ), and after 12 h, their thicknesses were 5 and 14 μm, respectively (Figure ). On the other hand, when pure Ti was nitrided at 1, 100°C for 12 h in atmospheric nitrogen, a 20-μm-thick TiN layer and a 50-μm-thick α-Ti(N) layer were formed [
4]. Also, when pure Ti was nitrided at 1, 250°C for 5 h in atmospheric nitrogen, a 35-μm-thick TiN layer was formed [
6]. The thicknesses of the nitride layers were larger, and Ti
2N was not detected in Vojtěch et al. [
4] and Seahjani and Cimenoglu [
6]. This discrepancy from the results obtained in this study is attributed to the higher nitriding temperatures and pressures employed in Vojtěch et al. [
4] and Seahjani and Cimenoglu [
6]. Intergranular β-Ti islands within the Ti(N) zone were not recognizable in the cross-sectional images because nitrogen is a potent α stabilizer, and the diffused nitrogen transformed the intergranular β into α. As the nitriding time increased, the thickness of the Ti-N layer and moreover, that of the Ti(N) zone increased together with the grain growth of the matrix.
Figure shows the TEM cross-sectional image of the Ti-6Al-4V alloy after nitriding at 850°C for 1 h. The elemental concentrations along spots 1 to 13 are listed in Table . It is however noted that the N-Kα, Ti-Lα, and V-Lα spectra overlap at approximately 0.39 keV, and the signal of nitrogen with a low atomic number is attenuated because of its low characteristic energy. Hence, the concentrations listed in Table are tentative. Nevertheless, nitrogen diffused interstitially to form an outer, 0.9-μm-thick Ti-N layer (spots 1 and 2). There should exist an inner Ti(N) zone below spot 3. Aluminum was locally segregated at spots 3 to 6 due to its limited solubility in the nitrides [
7].
| Table 1Concentration of spots 1 to 13 shown in Figure 4 (at.%) |
Figure shows the cross-sectional image of the Ti-6Al-4V alloy after nitriding at 850°C for 12 h. The elemental concentrations along spots 1 to 9 are listed in Table . Spots 1 to 6 correspond to the outer, 4-μm-thick Ti-N layer, below which the inner Ti(N) zone exists. Spots 3, 5, and 7 were determined to be TiN (Figure ), TiN plus α-Ti(N) (Figure ), and α-Ti(N) (Figure ), respectively. Aluminum tended to be depleted around the outer Ti-N layer and be enriched at spots 7 and 8. Such a tendency was however weak for vanadium.
| Table 2Concentration of spots 1 to 9 shown in Figure 5 (at.%) |
Figure shows the XRD patterns of the Ti-4Al-6V alloys after nitriding at 850°C for 1, 6, or 12 h, followed by oxidization at 700°C for 10 h in air. For the samples nitrided for 1 and 6 h, the initial TiN and Ti
2N nitrides were completely oxidized to rutile-TiO
2 on the α-Ti matrix. In the case of the sample nitrided for 12 h, Ti
2N with a preferred orientation along (002) was still retained underneath the TiO
2 surface scale. It is noted that Ti nitrides have better oxidation resistance than α-Ti because of the strong interaction between titanium and nitrogen, which decreases the thermodynamic activity of titanium and acts as a diffusion barrier against the inward diffusion of oxygen [
5].
The samples outlined in Figure were inspected by SEM as shown in Figure . All top views display fine oxide grains with rather rough surfaces. All of the cross-sectional images indicate the presence of 3- to 4-μm-thick oxide layers. In Figure , the Ti-N layers with original thicknesses of 0.8 to 2.3 μm were completely oxidized. In Figure , the Ti-N layer with an original thickness of 5 μm was partially oxidized.
3
