Influence of Processing Parameters on Laser-Assisted Reactive Sintering of a Mixture of Ni and Ti Powders

Abstract

Additive manufacturing (AM) plays a crucial role in the development of NiTi alloys, enabling the creation of complex and customized structures while optimizing properties for various biomedical and industrial applications. The aim of this paper was to investigate the influence of laser scanning speed on laser-assisted reactive sintering of a mixture of No and Ti powders. The samples were sintered at two different beam speeds, 4 and 5 4 mm/s and their morphological and microstructural characteristics were investigated. Scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM/EDX) and X-ray diffraction (XRD) analyses revealed the presence of intermetallic compounds rich in Ni and Ti for both scanning speeds; however, the scanning speed of 5 mm/s produced a microstructure with greater porosity, leading to a sintered body with poorer consolidation. Thus, employing a slower beam scanning of 4 mm/s seems to be a better alternative in the laser-assisted reactive sintering of NiTi alloys.

1. Introduction

Shape memory alloys (SMAs), such as NiTi, exhibit notable properties such as low rigidity, biocompatibility, and high corrosion resistance, making them ideal candidates for a variety of industrial applications, such as in the aerospace, biomedical, and automotive industries [1]. These applications are due to the distinctive features of shape memory effect (SME) and superelasticity (SE), stemming from a reversible martensitic transformation occurring at crystallographic and thermodynamic levels. These alloys consist of two phases: martensite (B19’), present at low temperatures, and austenite (B2), predominant at high temperatures [2,3].

SE is a crucial property of these alloys, manifested when tension is released after induced martensite formation within the austenite stability field under isothermal conditions. SME is the ability of the material to return to its original shape after deformation when heated above the transformation temperature to the austenitic phase. This mechanism imparts highly elastic behavior to the alloy. Transformation temperatures and mechanical properties of the alloy are affected by the presence of impurities [4,5]. Impurities such as TiN, TiC, and Ti₄Ni₂O oxides may be present in the material, especially in the case of fusion metallurgy processes such as vacuum induction melting (VIM) and vacuum arc remelting (VAR). Crucibles coated with zirconia or yttria are necessary to prevent contamination of the molten material, though this may result in oxide inclusions [6,7]. Despite the fact that the VAR process yields purer alloys, ensuring homogeneity remains a challenge, requiring repetitions [8,9].

Powder metallurgy (PM) is an alternative for the production of Ni-Ti alloys and other intermetallic compounds. However, conventional PM approaches face challenges in handling Ni and Ti powders due to their low compressibility and sinterability, necessitating techniques such as plasma sintering and hot isostatic pressing [10,11].

To overcome the challenges of fusion metallurgy and PM, reactive sintering (RS) emerges as a solution. This technique offers high homogeneity and purity through thermally activated and highly exothermic chemical reactions, known as self-propagating high-temperature synthesis (SHS), among powdered components [12,13,14]. The driving force for sintering is the reduction of surface free energy. This reduction may occur in two ways. One way is by reducing the total surface area due to the increase in the average particle size, known as “coarsening.” The other occurs through the elimination of solid–vapor interfaces and the creation of solid–solid interfaces in the grain boundary areas, which then grow and promote densification [15,16].

The technique involves powder mixing, cold pressing compaction, and subsequent sintering, with the possibility of obtaining a highly pure product by employing an appropriate protective atmosphere. The composition of the sintered material can be controlled by RS, with the final product structure influenced by parameters such as heating rate and initial particle size [17,18]. However, it is important to note that this process may result in porous materials due to various reasons such as volumetric changes in the crystal lattice, gas trapping, thermal migration, nonequilibrium diffusivity, Kirkendall effect, and residual porosity after pressing [19].

Previous studies analyzed the influence of specific parameters such as heating rate, reactive sintering temperature, and nickel and titanium particle size on microstructure, phase composition, and porosity [17,18,19,20]. The formation of exothermic intermetallic phases raises the system temperature, promoting homogenization. Due to its high tendency to porosity [21,22,23], RS can be an alternative to other powder metallurgy processes that have been used to produce self-lubricating materials or bone substitutes.

RS occurs over a wide range of temperatures. Intermetallic formation is highly exothermic, releasing heat that not only sustains but also propagates the reaction. Preferential formation of the Ti₂Ni second phase begins at 472 °C. At 632 °C, NiTi and Ni₃Ti formation occurs, followed by titanium transformation from alpha to beta phase at 890 °C and eutectic transformation at 942 °C, marking the transition to a self-propagating high-temperature synthesis (SHS) reaction [24].

The reactive sintering of NiTi alloys faces a significant gap in the available technical literature, particularly concerning critical parameters such as heating rates. Existing studies tend to focus on general aspects of microstructural behavior and the final properties of the material [18,19,24,25,26], without providing detailed information on the parameters necessary for process reproduction and optimization. This lack of data poses an obstacle to the standardization and continuous improvement of reactive sintering of NiTi.

This article aims to investigate the properties of NiTi alloys processed by reactive laser sintering with different scanning rates. Morphological/microstructural characteristics, phase formation, and thermal properties were analyzed using X-ray diffraction (XRD), scanning electron microscopy (SEM), energy-dispersive spectroscopy (EDX), and differential scanning calorimetry (DSC).

2. Materials and Methods

2.1. Raw Materials

The samples were prepared from an equiatomic mixture of high-purity powders of nickel and titanium (55.07 wt.% Ni and 44.93 wt.% Ti). The nickel powder was provided by JB Química (Indaial, Brazil), with 99.56 wt.% purity, and the titanium powder was produced by the hydride dehydrogenation (HDH) process from Ti bars Grade 1—with 99.5 wt.% purity, produced by the Instituto de Pesquisas Tecnológicas do Estado de São Paulo (IPT) (São Paulo, Brazil). The following steps were used to determine the composition of the powder mixture: (i) a green body pressed sample of the mixture was melted four times by an ArcMelter AM 200 vacuum arc remelting process (VAR), with a tungsten electrode and a water-cooled copper crucible, in Universidade Federal do Rio de Janeiro; and (ii) the casted sample was submitted for chemical analysis to Villares Metals by optical emission spectroscopy (OES), and O and C were determined by the LECO method. The results are shown in

Table 1. Chemical composition of the Ni and Ti powder mixture after casting in a VAR furnace (in wt.%).

Ni	Ti	Fe	O
55.2	44.3	0.102	0.290
Al	C	S	N
0.0148	0.091	0.015	0.0066

. Considering only the elements Ni and Ti, the powder mixture had, after melting, 55.48 wt.% Ni and 44.52 wt.% Ti (50.40 at.% Ni and 49.60 at.% Ti). This shows that the mixture had, approximately, the desired (equiatomic) composition.

2.2. Processing and Characterization of Ni-Ti Samples

Figure 1 presents the experimental scheme for preparing the samples. The powders were mixed in a roller mill for 10 min. After mixing, pellets were produced by cold uniaxial pressing, with 5 g of the Ni and Ti powder mixture (55.07 wt.% Ni and 44.93 wt.% Ti) and a compaction pressure of 55.5 MPa on a die with a diameter of 20 mm, followed by laser irradiation with an IPG YLR-2000 Yb:fiber laser of the Instituto de Estudos Avançados (IEAv), using a beam with a diameter of 2 mm, scanning di-directional/zigzag with a hatch distance of 2 mm, a power of 60 W, and scanning speeds of 5 and 4 mm/s (conditions 1 and 2), respectively, in an argon atmosphere, as shown in Figure 1.

After reactive sintering, the NiTi samples were cut transversely along the laser scanning lines. The samples were embedded and then ground along the diameter/thickness plane, and then polished with 300 mL of distilled water + 6 g of 1 μm alumina paste + 60 mL of hydrogen peroxide (H₂O₂). Subsequently, they were polished with 3 μm diamond paste with a solution of 300 mL of DP-Lubricant Red (Struers) + 30 mL of H₂O₂. Morphological characterization was conducted on the cross-sections of the samples, which were divided into three distinct regions. These regions were defined in relation to the thickness (t), diameter (d), and incidence face, referred to as incidence face (If) and opposite face (Of) to the laser. In addition to this identification, the samples were analyzed according to the regions of beginning (S), middle (M), and end (F), as illustrated in Figure 1.

Phase identification and peak positions were determined through X-ray diffraction using the PANalytical X’PERT PRO MRD diffractometer (Malvern Panalytical, Malvern, UK) with Co-Kα radiation, Fe filter, tube voltage of 45 kV, tube current of 40 mA, coupled θ/2θ, step of 0.050°, and scan range from 30 to 70°.

DSC analysis was performed to determine the phase transformations of the Ni-Ti mixture. A NETZSCH 404 F1 Pegasus calorimeter (NETZSCH Holding, Bavaria, Germany) was used with heating from room temperature of 20 °C to 1200 °C, at a heating rate of 10 °C/min, in an argon atmosphere, using an alumina crucible.

The microstructure of the samples was observed by an Olympus BX53M optical microscope (Olympus, Hamburg, Germany) and by a Quanta 250 FEG scanning electron microscope (SEM), using secondary electron (SE) and backscattered electron (BSE) detectors. The chemical composition was identified by an energy dispersive X-ray spectrometer (EDX). EDX/SEM analyses were carried out to examine the evolution of interdiffusion between Ni and Ti during the RS process under the conditions under study, using the green body as a basis for comparison.

3. Results and Discussion

3.1. Powder Characterization

The micrographs of the Ni, Ti, and powder mixture are presented in Figure 2. Ni powders, Figure 2a, had particles with irregular morphology [27].

The Ti HDH powder, illustrated in Figure 2b, had particles with irregular morphology, characteristic of the HDH process, which fragments the fragile metal hydride into powder before returning it to metallic form in the final stage of dehydration in vacuum [27,28].

In the micrograph of the powder mixture, presented in Figure 2c, it was observed that the Ni particles tended to aggregate on the surface of the Ti particles, increasing interaction between the powder particles. After uniaxial pressing, it was possible to observe in the green body small Ni particles surrounding the large Ti particles that filled most of the area observed in the micrograph in Figure 3.

EDX/SEM analysis (Figure 4) and scanning at different magnifications (

Table 2. Element concentration (in wt.%) by EDX/SEM of the Ni and Ti on the green body surface.

Magnification/Voltage	Ti (wt.%)	Ni (wt.%)	Ti (at.%)	Ni (at.%)
50×/30 keV	37.76	62.24	42.65	57.35
500×/30 keV	35.58	64.42	40.37	59.63
50×/20 keV	31.70	68.30	36.26	63.74
500×/20 keV	25.71	74.29	29.79	70.21

) highlighted much higher Ni concentrations than expected for the equiatomic composition. This was expected because the interaction volume of the electron beam was much smaller than the size of the Ni powder particles. Thus, depending on the particle size of the Ni powder, this made Ti undetectable when the Ni particles were on it (Figure 3a,b and Figure 4).

In Figure 5, the DSC thermogram of the powder mixture is presented, showing an onset temperature of transformation (T_s) of 581.67 °C, an exothermic peak temperature (T_p) of 722.29 °C, and an end temperature of transformation (T_f) of 739.48 °C. The temperatures were similar to the results obtained by Chen et al. [25], who reported that finer Ni particles along with coarser Ti particles tended to exhibit an exothermic peak in the temperature range found. Decomposition processes or the presence of stresses resulting from contraction may influence the transformation behavior, as well as the processes used to obtain the powders [26,29].

3.2. Microstructural and Morphological Characterization of Sintered Bodies

Microstructural analyses of Ni-Ti samples produced by laser-assisted reactive sintering revealed distinct patterns along the laser scan. In the higher magnification micrographs of the initial region, obtained via BSE and highlighted in Figure 6, a light gray tone indicative of advanced sintering near the upper laser incident surface was observed. This observation aligns with the findings of Zhao et al. [30], suggesting the influence of elemental diffusion gradient on microstructure formation. It was observed in the SEM images that condition 1 showed greater porosity than the sample of condition 2. The porosity was displayed as black spots, mainly in the initial region of the sample of condition 1, but it was also present in the initial region of condition 2. As the laser advanced along the sample, the energy density increased, resulting in a reduced porosity in regions (M) and (F), regardless of the scanning speed.

In BSE detection, darker regions were associated with elements of low atomic weight, while lighter regions were associated with elements of high atomic weight. Therefore, Ni or Ni-rich precipitates tended to appear in light regions, while Ti and Ti-rich precipitates were observed in dark regions. Additionally, it is important to note that the dark regions may have corresponded to rounded pores formed during the sintering process, being more evident in regions M and F. On the upper surface of the three regions, which was the area of laser incidence, the presence of pores was also observed. This suggests that the exothermicity of the ignition reaction reached a sufficient temperature to cause localized melting on the sample surface, resulting in gas retention due to increased solubility in the liquid phase [31,32].

Additionally, transition regions were identified in the intermediate region, where the growth of islands composed predominantly of Ti and Ti-rich phases was highlighted. These observations were corroborated by the tonality of the regions detected by BSE. In the final region, dark regions with faceted geometry in a homogeneous matrix were observed, attributed to the precipitation of Ti-rich phases. This can be explained by the influence of compositional and temperature ranges on the formation of precipitated phases within the matrix [33].

The analysis of sample densification along the laser scan revealed a significant improvement, highlighted by the increase in average pore size, especially in the initial region. This improvement in densification can be attributed to the high thermal input of the laser during its incidence on the samples. It is noteworthy that, during reactive sintering, the temperature inside the compacted powder particles increased rapidly, resulting in an internal pressure higher than the external one. This sudden pressure difference promoted an explosive wave, leading to sample expansion [24,34]. As a result, sintered samples often exhibited porosity, with pores of irregular geometries, contrasting with metals and alloys obtained by casting, which tended to have pores with rounded geometry due to trapped gases [13,33].

The SEM/EDX allowed the generation of compositional maps in the initial, intermediate, and final regions of the irradiated samples subjected to reactive sintering, as illustrated in Figure 7, and the percentual values obtained by EDX analysis are described in

Table 3. Element concentration (in wt.%) by EDX/SEM present in the reactive sintered NiTi bodies. Magnification of 50× and electron acceleration voltage of 20 keV.

	Condition 1 (5 mm/s)
Regions	Ti (wt.%)	Ni (wt.%)	Ti (at.%)	Ni (at.%)
Start	54.47	45.53	59.45	40.55
Middle	50.77	49.23	55.84	44.16
End	50.74	49.26	55.80	44.20
	Condition 2 (4 mm/s)
	Ti (wt.%)	Ni (wt.%)	Ti (at.%)	Ni (at.%)
Start	59.17	40.83	63.98	36.02
Middle	53.38	46.62	58.39	41.61
End	49.84	50.16	54.92	45.08

.

Table 3 shows the elemental composition by EDX/SEM for the samples with a scanning speed of 5 mm/s (Condition 1) and 4 mm/s (Condition 2), indicating an insufficient thermal input for compositional homogenization of the entire three regions. However, it is necessary to compare these results with those of the green body, Table 1, the physical principles of the EDX/SEM analysis, and its limitation regarding a two-dimensional surface analysis associated with a low penetration of the electron beam to generate the X-ray signal to be detected by EDX. With this point of view, it was possible to infer the evolution of the interdiffusion phenomenon between Ni and Ti from the neighboring powders of the mixture in the sintered bodies result of the RS process. It was known that both Ni and Ti would be detected in the alloy and that this now occupied part of the area where there was only Ni or Ti on the surface. Therefore, when observing the decrease in the Ni content (compared with the green body), along the regions of the sintered material, there was evidence that there was a thickening of the alloy layer in the start and middle regions, The closer the detected values of Ni and Ti were, the greater the evidence of the presence of melting in the region, and, in fact, the existence of casting product in the middle (partially) and end (completely) regions, as shown in Figure 6, were more significant in condition 2 (4 mm/s) due to a greater density of applied energy.

The formation of Ti-rich intermetallic compounds in the end region mainly may have been attributed to the difference in the melting points of the two elements (Ni: 1453 °C and Ti: 1668 °C). Additionally, the larger particle size and faceted geometry of Ti might have hindered complete interdiffusion between the elements, preventing the uniform formation of the alloy. The difference in melting points led to an increase in the temperature of the molten material, which consequently may have reduced the Ni atom content through vaporization [31].

These results were consistent with observations obtained by XRD, which also indicated the formation of intermetallics. It is noteworthy that the major Ti and Ni regions suggested a lack of powder coalescence during the reactive sintering process. Furthermore, the presence of these distinct regions indicated that the thermal input provided during the process was not sufficient to achieve compositional homogenization along the entire region.

XRD patterns are observed in Figure 8. Alongside the results presented by SEM and EDS, the presence of Ni-rich intermetallics was observed, such as Ni₃Ti, Ni₃Ti₂, and Ni₄Ti₃, and Ti-rich intermetallics, such as Ti₂Ni, consistent with the compositional fluctuations observed by EDX and the findings of Novák et al. [19], where similar phases were identified, such as the formation of Ti₂Ni, indicating that the transformation during the self-sustaining process initiated with the Ti allotropic transformation, followed by the reaction with Ni, forming Ni₃Ti, phases which, after decomposition, formed the eutectoid intermetallic NiTi. Furthermore, metallic Ti peaks were also evident in EDX/SEM analyses in the start and middle regions of the sintered bodies (Figure 7). It was also considered the occurrence of oxidation—mainly from titanium powder before synthesis provided by RS (in Ti₂O)—because the process was not conducted in a vacuum or chamber with an inert atmosphere with a low concentration of oxygen, instead of Ti₄Ni₂O (which presented its diffraction peaks very close to Ti₂Ni) or the reasons highlighted above for the Ti₂Ni. Finally, it was impossible to conclude that the material presented phase B2, the austenitic phase associated with the equiatomic intermetallic NiTi, and it was also impossible to confirm that at a position close to 49.67° one of the peaks was associated with one of these phases: (110)B2, (20-2)R, (220)R, and/or (201)Ni3Ti. This way, this peak was marked as undefined with symbol (*).

The presence of TiO_x oxide is associated with solid-state oxidation due to the reactivity of Ti with oxygen [6,7]. The results obtained in this study are further justified by the utilization of finer Ti particles (below 150 μm), leading to the undesirable microstructure composed of NiTi₂, where the surface interaction between Ni and Ti particles can be observed, with some Ni grains situated on the surface of Ti grains, without penetration into the grain interiors [19,24].

The difference in melting points between Ni and Ti—1455 °C and 1668 °C, respectively—plays a fundamental role in the temperature at which the alloy forms, as well as affecting the size and geometry of Ti particles. This dissimilarity results in the absence of complete interdiffusion between the elements during alloy formation, leading to the formation of regions with higher Ti concentration [19,24,31]. Ti tends to diffuse towards the edges of molten Ni regions, while the reverse, i.e., Ni diffusion towards solid Ti, does not occur readily. Even though the process is designed to circumvent this diffusion step through rapid heating, particle size still significantly affects this phase of the reaction mechanism [24,30,32].

4. Conclusions

In this paper, the influence of laser scanning speed on the reactive sintering process of a Ni-Ti alloy, and the morphological and microstructural characteristics of the alloy, was investigated. After processing, the samples were separated and characterized, both in powder form and in sintered bodies, which were cut and evaluated in different regions to analyze the laser impact on the microstructure. Based on the experimental results, we inferred that the mixing of Ni-Ti powders was possible due to the tendency of Ni to agglomerate on Ti particles.

Thermal analysis revealed the transformation temperatures of the powders, which were consistent with the literature on the phase transformation of Ni-Ti alloys for additive manufacturing. Morphological analyses showed the presence of a series of intermetallic phases, both rich in Ni and rich in Ti, formed during reactive sintering. Additionally, we observed that a higher beam speed (condition 1) resulted in greater porosity in the samples. The formation of these intermetallics was confirmed by XRD analysis. In summary, we were able to sinter a Ni-Ti alloy using both speeds, but the lower speed (condition 2) yielded a sample with better consolidation since it exhibited lower porosity after processing.