Depth profiling analysis of the nitriding layer formed by gas nitriding of Ti13Nb13Zr alloy

The gas nitriding of Ti13Nb13Zr alloy was performed in a pure nitrogen atmosphere at 1200 °C with different holding times. The variations in the phase constituents, microstructure, and mechanical properties of the nitriding layer with depth were characterized by SEM, XRD, and nanoindentation. Depending on the phase distributions in depth, the nitriding layer could be sequentially divided into an external nitride layer mainly consisting of TiN, an internal nitride layer consisting of TiN, Ti2N and TiN0.3, and an N diffusion region consisting of TiN0.3. The mechanical properties were closely associated with the phase constituents, where the nanohardness of the internal nitride layer, internal nitride layer N diffusion region and substrate were about 17.126±0.399, 12.120±0.386, 5.627±1.080, and 3.632±0.116 GPa, respectively. In the nitriding process, the N diffusion region containing needle-like TiN0.3 precipitates formed in the initial nitriding stage, and then the precipitates were gradually converted to the external and internal nitride layers, whose thickness increased with the nitriding time. Furthermore, the influence of the alloying element redistribution on the N diffusion mechanism was discussed.


Introduction
Ti13Nb13Zr alloy is an attractive material for hip prosthesis bearing owing to its high specific strength, low Young's modulus, excellent corrosion resistance and good biocompatibility [1].Nevertheless, its low hardness, high coefficient of friction and poor wear resistance restrict its application in the field of hip prosthesis bearings [2].One effective solution is surface nitriding treatment, which can combine the superior tribological properties of the nitriding layer and the excellent bulk properties of titanium alloy.To develop nitriding layers with excellent comprehensive properties, numerous surface nitriding techniques such as physical vapor deposition (PVD), chemical vapor deposition (CVD), plasma, and thermochemistry have been investigated [3][4][5][6].Among them, gas nitriding process seems to be a promising strategy for applying titanium alloy to hip prosthesis bearings due to its deep nitrogen penetration, high adherence to the substrate and ability to handle complexshaped parts.
Previous studies on the gas nitriding treatments of titanium alloys have covered various types of titanium alloys with different chemical compositions and investigated the effect of varied treatment parameters on the microstructure, phase transformation behaviors, growth kinetics, and surface properties of the nitriding layers [7][8][9].In general, titanium alloy can obtain an outmost nitride layer consisting of TiN/Ti 2 N and an N diffusion layer consisting of TiN 0.3 after gas nitriding treatment, whose phase constituents, microstructure and thickness have significant impacts on the wear and corrosion resistance as well as biocompatibility.However, the research on the gas nitriding treatments of Ti13Nb13Zr alloy remained insufficient, and the relevant reports mainly focused on the microstructural development and the characterization of the required surface properties [10], while did not clarify the nitriding mechanism and the exact relation between the elemental composition, phase constituent and property evolution in the nitriding layer.In fact, these played crucial roles in determining nitriding parameters to achieve optimum properties of nitriding layer.
In this work, gas nitriding at 1200 °C with different holding times was conducted to investigate the gas nitriding mechanism of Ti13Nb13Zr alloy.In addition, the depth profiling characterization of the nitriding layer was studied systematically via SEM, XRD and nanoindentation, which could provide practical guidance on selecting nitriding parameters for achieving nitriding layers with optimum performance.

Material and sample preparation
The Ti13Nb13Zr alloy was purchased from Peshing New Metal (ChangZhou) Co., LTD, and its chemical composition is presented in table 1.The sample used for gas nitriding treatment was prepared with a size of 10 × 10 × 5 mm using wire cutting.Prior to nitriding, the samples were ground using SiC sandpaper from 180 # to 3000 # and then polished with silica solution (they were provided by DongGuan Songyoo Electronics Co., Ltd).Afterwards, the samples were ultrasonically cleaned in acetone for 10 min.

Gas nitriding
Gas nitriding treatment for Ti13Nb13Zr alloy was conducted in a high-purity (99.999%) nitrogen gas atmosphere using a tube furnace (GSL-1400X, HF-kejing, China).After placing the sample, the tube was vacuumed down to 1 Pa by a vacuum system (GZK 103D, HF-kejing, China) and then filled with nitrogen to atmospheric pressure.Afterwards, the Ti13Nb13Zr alloy samples were nitrided at 1200 °C and held for 0, 2, 4, 6, 8, and 10 h, respectively, as shown in figure 1.For all the nitriding samples, the heating and cooling rates were set at 5 °C /min, and the nitrogen flow rate was maintained at a constant level of 100 ml min −1 .

Characterization
The phase identification of the nitriding layer was examined by X-ray diffraction (XRD, D8 Advance, Bruker, Germany) at a voltage of 40 kV and a current of 30 mA with Cu Kα radiation in the 2θ range of 20°-80°.The chemical state was analyzed by X-ray photoelectron spectroscopy (XPS, EscaLab 250Xi, Thermo Scientific, USA) using a monochromatic Al Kα radiation at 1486.6 eV, and the C 1 s peak of 284.8 eV was used for energy calibration.The cross-sectional microstructure and element composition of the nitriding layers were characterized by Scanning Electron Microscope (SEM, SU3500, Hitachi, Japan) equipped with energy dispersive spectroscopy (EDS).The mechanical properties of the nitriding layer were measured using nanoindentation  instrument (Tribo Indenter, Hysitron, USA) with a diamond Berkovich indenter.The maximum load of 2000 μN was applied for 5 s during the indentation, and the loading and unloading times were set to 5 s, respectively.

Phase identification
Figure 2 shows the XRD patterns obtained from the top surfaces of the untreated Ti13Nb13Zr alloy and the nitriding samples with different holding times.It can be seen that the untreated Ti13Nb13Zr alloy was composed of α-Ti (ICDD 04-003-5042) and β-Ti (ICDD 97-004-4391).For the nitriding samples, the peak positions at 36.60°(111), 42.44°(200), 61.58°(220), 73.80°(311), and 77.73°(222) belonged to TiN phase (ICDD 04-002-5535), which was the dominant phase of the nitriding samples regardless of the holding time.It should be noted that TiN 0.3 (ICDD 00-041-1352) and Ti 2 N (ICDD 01-076-0198) were present at high intensities in the S0 and S2 samples, whereas their amount decreased with the increasing holding time and finally disappeared in the S8 sample.Although XRD characterization was not quantitative, the changes in the relative intensities of the peaks could reflect the changes in relative proportions of the different phases.Thus, the gradually decreased relative intensities of the TiN 0.3 and Ti 2 N peaks indicated that a phase transformation from the TiN 0.3 and Ti 2 N phases to the TiN phase occurred during the nitriding process.Additionally, some TiO 2 was detected in nitriding samples, which can be attributed to the small amount of oxygen remaining in the tube furnace.
The chemical state of the nitriding layers can be determined by XPS measurements.Figure 3 exhibits the XPS survey and Ti 2p, N 1 s, and O 1 s core level spectra of the S6 sample.As shown in figure 3(a), the XPS survey confirms the presence of Ti, N and O elements in the nitriding layer.Since the orbital and spin motions of the electrons in atoms, the Ti 2p peak has significantly split spin-orbit components, which are Ti 2p3/2 and Ti 2p1/ 2, and the splitting Δ-value generally varies with the chemical state (Δmetal = 6.1 eV, Δoxide = 5.7 eV, Δnitride = 6.0 eV).In figure 3(b), the Ti 2p spectrum was deconvoluted into eight synthetic peaks.The peaks at the binding energies of (458.77,464.47 eV) and (454.79,460.79 eV) were attributed to the Ti-O bond in the TiO 2 and the Ti-N bond in the TiN, respectively, and the peaks located at the binding energies of 457.86 and 463.26 eV represented the satellite feature of the TiN, whose intensities would be affected by TiN x stoichiometry [11].According to reports in the literature [12], TiN x O y compounds could be formed by the reaction of TiN with O. Thus, the peaks at the binding energies of 456.50 and 462.05 eV were believed to come from the Ti-N-O bond in the TiN x O y compounds.Notably, some reports considered that the characteristic binding energies range of the TiN x O y compounds might overlap that of the TiN satellite, and the peaks of the TiN x O y compounds and satellite were sometimes fitted into a single peak [13][14][15].In N 1 s core-level spectrum (figure 3(c)), the peaks at the binding energies of 397.18 and 394.60 eV corresponded to the N-Ti bond in the TiN and the N-O bond in the TiN x O y , respectively, while the peak at 399.73 eV could be interpreted as the satellite feature of the TiN.The O 1 s core level spectrum in figure 3(d) was decomposed into two characteristic peaks.The peak located at the

Depth profiling characterization
In order to investigate the variations in phase constituents and mechanical properties with depth, XRD and nanoindentation tests were simultaneously performed at different depths.Figure 4 shows the analytical results of the S6 sample.In accordance with the microstructural evolution with depth provided by the cross-sectional image (figure 4(a)), the nitriding layer can be roughly divided into three different regions from top to bottom, and A-A′ (5 um from the top surface), B-B′ (20 um from the top surface) and C-C′ (60 um from the top surface) were representative planes located inside of the Region I, Region II, and Region III respectively, which were obtained by gently polishing the external material.
Figure 4(b) presents the XRD patterns recorded from the three planes at different depths.The XRD pattern taken from the A-A′ indicated that the TiN was the primary phase of Region I, while the Ti 2 N and TiN 0.3 were the minor ones.Notably, the TiO 2 phase did not appear at this place, suggesting that oxides only formed on the outermost surface.For Region II, the XRD pattern obtained from the B-B′ illustrated that the TiN was still the primary phase, however, the diffraction peaks of the Ti 2 N and TiN 0.3 phases became stronger.After completely polishing off Region I and II, the XRD pattern of the C-C′ exhibited that Region III consisted of the TiN 0.3 phase, whilst some β-Ti phase began to appear.Depending on the phase distribution with depth, the three regions could be identified as Figure 4(c) shows the load-displacement curves obtained from the A-A', B-B', C-C' and substrate (the core position of the nitriding sample).As can be seen, the maximum indentation depth of the substrate was considerably larger than those in the nitriding layer, indicating that the nitriding treatment significantly improved the nanohardness of Ti13Nb13Zr alloy.The detailed mechanical properties parameters extracted from the load-displacement curves are listed in table 2. The nanohardness of the A-A′, B-B′, C-C′ and substrate were 17.126±0.399,12.120±0.386,5.627±1.080,and 3.632±0.116GPa, respectively.This decreasing trend coincided with the top-to-bottom phase distribution shown in the XRD analysis.In addition, it was found that the nanohardness values of the A-A′, B-B′, C-C′ and substrate were approximately consistent with those of the corresponding positions where they were located in the cross-sectional nanohardness profiles shown in figure 4(d).Although there were some slight deviations due to the insufficiently fine control in the thickness of the removed material during the polishing process, this result could still suggest that the whole nitriding layer was compact and uniform.In general, two principal H to E ratios, viz., H/E and H 3 /E 2 (H being the hardness and E the elastic modulus), can also be used to estimate the mechanical properties of the nitriding layer.H/E characterises the resistance of the material to elastic deformation, and H 3 /E 2 represents the resistance to plastic deformation [16].It can be seen that the H/E and H 3 /E 2 ratios obtained from the nitriding layer were significantly higher than that of the substrate, indicating gas nitriding can effectively improve the resistance of Ti13Nb1Zr alloy to elastic and plastic deformation, which was beneficial for improving the wear resistance.Additionally, except for the H/E value of B-B', two ratios gradually decreased with the depth, revealing a gradient decrease in the mechanical properties of the nitriding layer, which was related to the phase distribution.

Nitriding mechanism
Figure 5 shows the cross-sectional microstructures of the nitriding layers formed under the different holding times, which can provide essential information to investigate the nitriding mechanism of Ti13Nb13Zr alloy.As  mentioned, the nitriding layer comprised the nitride layers (external and internal) and the N diffusion region.It can be seen that all the nitride layers (figures 5(a)-(f)) were compact, homogeneous, and well adhered to the substrates.As the holding time extended, the nitride layer gradually developed inwards to the substrate, and the thickness of the external nitride layer of S0, S2, S4, S6, S8, S10 samples were about 2, 4, 5, 6, 8 and 10 μm, respectively, which caused an increasing surface hardness as shown in figure 6.In particular, the surface hardness of S10 sample reached 1302 HV, which was about 5.3 times that of the bare Ti13Nb13Zr alloy (244 HV).Due to the uneven growth fronts, the exact thickness of the internal nitride layers could not be obtained.However, it was certain that the overall thickness of the nitride layer of the sample nitrided at 1200 °C for 10 h eventually exceeded 50 μm.
Concerning the N diffusion region, the TiN 0.3 precipitates did not generate a continuous layer as observed for the gas nitriding pure titanium [11], while distributed in the form of coarser needles and seemed to have strong orientation relationships with the substrate.Definitely, alloying elements played a crucial role in the final morphology evolution of the precipitates.According to the report by Buscaglia et al [17], initial nucleation of the TiN 0.3 precipitates in the nitriding Nb-Ti alloys preferentially occurred at grain boundaries and impurity sites rather than homogeneous nucleation, and subsequent internal nitridation mainly occurred by enlarging the existing precipitates.In this process, the Nb element was squeezed out from the TiN 0.3 precipitates owing to its β-stabilizing action and limited solubility in the TiN 0.3 phase, leading to the Ti segregation in the precipitates and Nb enrichment in the substrate, which ultimately resulted in opposite Nb and Ti concentration gradients from the substrate to surface.Since the growth rate of the precipitates was determined by the rate at which the precipitate-forming element reached the precipitation site, the final morphology evolution of the precipitates was strongly influenced by the rapid diffusion of N and the slow Nb-Ti interdiffusion in the substrate [17], which may be the main reason of the formation of the coarse needle-like TiN 0.3 precipitates having strong orientation relationships with the substrate.
Generally, the morphology of the precipitates was primarily established at an early nitridation stage attributed to the maximum inward nitrogen flux.Afterwards, the external TiN 0.3 precipitates were progressively converted to nitride composed of Ti 2 N/TiN as the nitriding time increased and eventually evolved into the external and internal nitride layers.Notably, some short dots and long laths were observed to distribute discontinuously along the boundary between the external and internal nitride layers, and this situation was even more pronounced for the samples that underwent longer nitriding times (figures 5(e) and (f)).Figure 7 shows the SEM image of the nitriding layer of the S8 sample, together with the corresponding EDS mapping and line scanning profile of Ti, Zr, Nb, and N elements.As can be seen, the lath displayed a distinctly different chemical composition from that of the nitride layer, and it was enriched with Nb (with some content of Zr) and depleted in Ti and N, which was similar to the Nb segregation below the nitride layer [18].According to the computation of the standard free energy reported in [19], the nitriding treatment at 1200 °C was not beneficial to generate the nitride of Nb compared with titanium nitride.In addition, the EDS point measurement exhibited that the composition of the lath was 63.96 Ti Wt%, 31.98 Nb Wt% and 5.06 Zr wt%, while N was not detected, indicating that Nb was indeed not involved in the formation of a nitride.Therefore, it can be deduced that the Nb in these laths was squeezed from the generated external nitride layer and formed local segregation.
Gas nitriding of titanium alloy is a complex process, and the influence of various parameters involving temperature, nitriding time and nitrogen partial pressure were studied.In addition, the effect of alloying elements and their redistribution on N diffusion during the nitriding process has been demonstrated [20], which could alter the N diffusion path and thereby influence the phase constituent.Figure 8 displays the main mechanisms involved in the gas nitriding of Ti13Nb13Zr alloy in 1200 °C, which can be summarized as follows: In the initial stage of the nitriding process, N accumulated on the sample surface and diffused into the substrate.As the dissolved N in the Ti13Nb13Zr alloy reached saturation, TiN 0.3 precipitates began to nucleate at grain boundaries and impurity sites, and the subsequent nitridation developed through the growth of the existing precipitate instead of the formation of new precipitates, eventually leading to the formation of the needle-like morphology.As the holding time increased, the N concentration of the precipitates close to the surface gradually exceeded the N solubility in the TiN 0.3 phase.Thus, these precipitates were progressively converted to Ti 2 N and then to TiN with continuous N dissolution.In this process, Nb was rejected from the N-enriched regions to the surrounding Ti substrate, and formed segregation below the external nitride layer and the internal nitride layer.Considering that Nb worked as a barrier in N diffusion, uniform inward growth of the nitride layer cannot be applied in the gas nitriding of Ti13Nb13Zr alloy.Generally, the Nb segregation along the boundary between the external and internal nitride layers occurred in short dots or long laths.Since the N diffusion channels flowing through these regions were blocked, the underneath N fluxes were reduced, which   would not allow the TiN0.3 below the segregation to be converted to nitride as successfully as the adjacent regions.In consequence, the phase distribution of the internal nitride layer would not be homogeneous in the lateral direction, and the TiN, Ti2N, and TiN0.3 may be present at the same depth of the internal nitride layer, which was consistent with the XRD result obtained from the depth of 20 μm.In addition, the heterogeneity over the thickness (jagged morphology) of the internal nitride layer could also be interpreted by the uniform Nb segregation below the internal nitride layer.

Conclusions
The Ti13Nb13Zr alloy was gas nitrided in a pure nitrogen atmosphere at 1200 °C with different holding times, and the nitriding layers were characterized to capture the information of the variations in chemical composition, phase constituent, microstructure and mechanical properties with depth.Based on the results obtained, the following conclusions can be drawn: (1) The TiN, Ti 2 N, and TiN 0.3 phases were formed after the gas nitriding of Ti13Nb13Zr alloy at 1200 °C.A few TiO 2 and TiN x O y phases were also formed due to the residual oxygen, but only present on the outermost surface.
(2) Depending on the phase distributions in depth, the nitriding layer could be sequentially divided into an external nitride layer mainly consisting of TiN, an internal nitride layer consisting of TiN, Ti 2 N and TiN 0.3 , and a N diffusion region consisting of TiN 0.3 .The mechanical properties of the nitriding layer closely followed the phase distribution.
(3) Alloying elements redistribution took place during the nitriding process, which resulted in the formation of the needle-like TiN 0.3 precipitates with strong orientation relationships with the Ti substrate.In addition, the localized Nb segregation along the boundary between the external and internal nitride layer blocked some N diffusion channels, leading to inconsistent downward N diffusion and ultimately resulting in the appearance of multiple phases in the internal nitride layer.

Figure 1 .
Figure 1.Process flow diagram of the gas nitriding treatments of Ti13Nb13Zr alloy with different holding times.

Figure 2 .
Figure 2. XRD patterns of the untreated and nitriding Ti13Nb13Zr samples.

Figure 4 .
Figure 4. Depth-evolution of phase constituent and mechanical properties of the nitriding layer of the S6 sample (a) cross-sectional SEM image; (b) XRD patterns taken from the different depths; (c) load-displacement curves taken from the different depths; (d) crosssectional nanohardness profile.

Figure 6 .
Figure 6.Surface hardness of the untreated and nitriding Ti13Nb13Zr samples with different holding times.

Figure 7 .
Figure 7. Cross-sectional SEM images and EDS analysis of the S8 sample.

Figure 8 .
Figure 8. Schematic representation of the gas nitriding of Ti13Nb13Zr alloy in 1200 °C.

Table 1 .
Composition of the applied Ti13Nb13Zr alloy in mass percent.

Table 2 .
Mechanical properties parameters extracted from the load-displacement curves.