1. Introduction
Nickel–Titanium (NiTi or Nitinol) shape memory alloys have been employed in numerous engineering sectors during the past few decades due to their superior shape memory effect, biocompatibility, and super elasticity characteristics [
1,
2]. NiTi alloys have been widely employed in biomedical, hydrospace, automotive, aerospace, and medical industries [
3]. One of the most widely used industries for NiTi is biomedical engineering, particularly in fabricating orthodontic arc wires for cardiovascular stents, surgical tools, miniature devices, or orthopedic implants [
4]. With regard to the mechanical integrity of the welded joints, fusion-based joining techniques such as gas tungsten arc welding [
5] and laser-welding [
6] have revealed the best outcomes. Laser-welding is the most-used welding technique over other fusion-based welding processes due to its characteristics, such as a reduced heat-affected region, deep and narrow fusion zone, better production rate, lower residual stress, higher energy density, and excellent control of process variables [
7,
8]. Laser-welding is a highly useful technique in aerospace [
9,
10], transportation [
11,
12], and biological [
13,
14,
15] contexts because it can generate high-quality welds with few errors and is adaptable to many materials and complex geometries. During the welding, NiTi alloys may undergo a shift in the chemical composition, resulting in the transformation temperatures of thermally altered areas; i.e., after welding, the microstructure of the parent NiTi alloy with a fully austenitic structure may exhibit martensite or a combination of martensite and austenite in areas exposed to welding. These microstructural changes can significantly affect the mechanical integrity and corrosion performance of welded NiTi structures. Due to the poor formability and machinability characteristics, NiTi alloys are often limited in manufacturing small and miniature devices. Many studies have evaluated the mechanical integrity and corrosion performance of wrought NiTi alloys in different corrosive environments, while few studies have reported on the welded NiTi structures.
Oliveira et al. [
1] studied the influence of laser-welding process variables on the fraction of austenite and martensite phases in NiTi alloys. The fraction of martensite varied from the base metal (BM) to the weld metal (WM) due to the higher thermal gradients during welding. Mehrpouya et al. [
6] investigated the microstructural features, shape memory effect characteristics, and hardness of NiTi sheets via a diode laser. Epitaxial growth was noticed in the fusion zone, and the formation of precipitates such as NiTi
2, Ni
3Ti, and Ni
4Ti
3 was confirmed. Also, the microhardness of the welded NiTi alloy at the weld center (270–300 HV) showed a decrease in hardness compared to the BM (320 HV). Datta et al. [
16] welded 1 mm NiTi sheets with a Yb-doped fiber laser system and explored the correlation between input–output responses using different artificial neural network (ANN) models. The output responses, such as the weld bead characteristics, hardness, and tensile properties, were evaluated. In addition, the microstructure varied from the BM to the WM, and a drastic drop in the tensile properties was observed, while the Bonobo optimizer showed less deviation in predicted values than that of other ANN models. Kannan et al. [
17] evaluated the corrosion behavior of laser-welded NiTi sheets in a 0.9% NaCl solution at 37.5 °C. The corrosion behavior of the NiTi WM was better than that of the BM due to an increase in the Ti/Ni ratio. Prabu et al. [
18] explored the functional and corrosion properties of friction-stir-welded NiTi sheets having 1.2 mm thickness. The average tensile strength and elongation of the NiTi WM was 605 MPa and 7% in comparison to the BM (990 MPa and 27%). Also, the localized corrosion behavior of the NiTi WM was reduced due to the microstructural difference across the weld, resulting in the formation of nonuniform oxide layers.
Biocompatibility of the NiTi alloys has been assessed under different simulated environments [
19,
20]. The release of Ni ions is the dominant factor that controls the compatibility and safe use of NiTi alloys in the human body. Concerning the present study, Sevilla et al. [
21] examined the phase transformation, mechanical integrity, and corrosion characteristics of NiTi orthodontic archwires using Nd: YAG laser-welding for selective force application. After welding, the microstructure of NiTi did not change considerably. At the same time, tensile properties were reduced significantly, and the release of Ni ions in the artificial saliva medium was within the biological tolerance range. Toker et al. [
4] studied the microstructural characteristics and localized corrosion performance of wrought NiTi alloys in a simulated body fluid (SBF) environment. The corrosion results indicated the formation of stoichiometric TiO
2 oxides near high-energy zones like dislocation networks. Also, Ni- and Ti-rich intermetallic phases existed in the corroded surfaces and were attributed to the higher inner diffusion, followed by a Ni release. Kassab et al. [
22] assessed the fracture of NiTi wires due to corrosion in simulated oral environments. The NiTi wires demonstrated sufficient resistance to corrosion in conditions with 9 g/L NaCl and SBFs, like saliva. In addition, the fracture type was brittle, while the corrosion pits acted as crack nucleation points. The corrosion resistance of NiTi alloys mainly depends on the passive film formation on the surface. However, the release of ions can significantly influence the life of bioimplants. Furthermore, the passive film may deteriorate throughout its service life. The physiological environment can influence the NiTi alloy’s corrosion performance and oxide products. The welding of NiTi alloys can alter the microstructure and intermetallic formation, which can influence the release of metal ions into the body, causing health issues. Beyond that, the existence of martensite influences the dislocation formation at the austenite and martensite phase boundaries, resulting in enhanced selective corrosion attacks. Most studies have reported optimizing process variables, microstructures, and mechanical properties of welded NiTi alloys. In this study, a novel attempt is made to evaluate the microstructural features, mechanical properties, and corrosion performance of laser-welded NiTi sheets. Electrochemical tests were performed in SBF, and a comparison was made between un-welded and laser-welded NiTi specimens. The interaction between the laser-welding process and the resultant alloy behavior is uniquely explored in this study, offering new insights into the material’s appropriateness for biomedical applications, especially concerning long-term implant durability and performance.
2. Materials and Methods
Annealed and oxide-free sheets of NiTi with a 1 mm thickness were used in the present study. The nominal composition of the NiTi sheets conforms to the elemental limits in accordance with ASTM F2063-18 standards [
23], as presented in
Table 1. Before welding, the coupons of NiTi were mechanically scrubbed with a stainless steel wire brush and then cleaned with acetone to remove the contaminants from the specimen surface. The power source for the welding process was a Yb-YAG fiber laser-welding system (Make: TRUMPF GmbH, Ditzingen, Germany) with 4 kW power, a 0.2 mm focal spot size, and a 1.03 μm wavelength. From iterative trial and error runs, the optimized set of process variables was selected to fabricate the coupons to a butt joint configuration. The coupons for welding were rigidly clamped without any gap. From existing literature, the criteria for selecting the processing variables to fabricate the joints were based on the weld bead characteristics, i.e., the weld bead with full penetration depth. The range for conducting trials was considered from an existing study [
24], and the process variables were slightly modified to obtain better results in the present study. The process variables, such as the beam power (W) and welding speed (mm/min), were controlled in the range of 850 to 950 W and 2000 to 2500 mm/min, respectively. The optimized set of process variables is presented in
Table 2. The top and bottom sides of the blanks to be welded were supplied with 99.99% pure industrial-grade argon at a constant flow rate of 15 L/min to prevent contamination in welds.
After welding, the weld bead region with a uniform cross-section was considered for specimen preparation using an electric discharge machining (EDM) process. The microstructural features and elemental distribution in NiTi’s BM and WM specimens were examined with a Zeiss (Zeiss, Oberkochen, Germany) field-emission scanning electron microscope (SEM) equipped with energy-dispersive X-ray spectroscopy (EDS). For microstructural analysis, the transverse section of the laser-welded NiTi was prepared using an EDM and followed standard procedures, as mentioned in the ASTM E3-11 (2017) standard [
25]. The WM sample was mirror-polished and etched with a HNO
3 + H
2O + HF solution in the ratio of 4:5:1. X-ray diffraction (XRD) was used to determine the phases in the NiTi WM. A Pan Analytical Empyrean diffractometer (Malvern Panalytical GmbH, Kassel, Germany) measured the 2θ range of 20–100 degrees. It was equipped with CuKα radiation with a wavelength of 0.1546 nm. The XRD scanning was performed at 0.05 degrees per second with a step angle of 0.02 degrees. The hardness measurements were performed by following the ASTM E384-22 standard [
26] using a Struers Duramin-4 Vickers Hardness Tester with an applied load of 500 g and a dwell time of 15 s. The position for the hardness measurement was chosen to be 200 mm below the top and bottom surfaces of the weld bead, with 500 µm and 100 µm, respectively, between subsequent indentations in the x and y axes. According to the ASTM E8/E8M-22 standard [
27], tensile specimens of the NiTi BM and WM were prepared and tested using an 8801 servo hydraulic universal tensile testing machine (Make: Instron, Norwood, MA, USA) with a cross head speed of 1 mm/min. Three specimens from the NiTi BM and WM were tested to evaluate the average tensile properties. Potentiodynamic polarization testing in SBF at 37.5 °C, following the ASTM G61-86 (2018) standard [
28], was used to investigate the localized corrosion performance of the NiTi BM and WM. Tafel plots were obtained using a classical three-electrode system (Make: ACM Hill AC Instruments, Grange-Over-Sands, UK). The composition and pH of the SBF are shown in
Table 3. The corrosion experiments were performed three times to ascertain the average corrosion properties. Equation (1) was utilized to compute the corrosion rate [
29].
where I
corr is the corrosion current density in mA/cm
2, E
w is the material’s equivalent weight in g, and ρ is the material’s density in g/cm
3.
The EIS measurements were also performed with a 10 mV RMS sinusoidal perturbation with reference to OCP within the frequency limit of 105 Hz to 102 Hz. Using the Zview software, version 3.2, EIS spectrums from the experiments were matched. The corrosion pits were analyzed using SEM and EDS techniques to analyze the morphology and distribution of various elements around the pits.
3. Results and Discussion
The macrostructure of the joint fabricated with optimized process parameters is shown in
Figure 1. The laser-welded joint was devoid of defects such as cracks and had a uniform shape with full penetration.
Figure 2 represents the SEM images of the laser-welded NiTi joint at different locations, specifically the BM, heat-affected zone (HAZ), and WM. Because of the higher temperature gradient experienced during laser-welding and the impact of solidification, the fusion zone (FZ) usually had a coarser grain structure than the HAZ. Due to the temperature difference at the center of FZ, the grains expanded, and the limited heat conductivity of NiTi alloys made it difficult for the heat to spread out relatively toward the candidate material.
Figure 2a shows the microstructure of the NiTi BM, dominantly austenitic with the B2 austenite phase. The FZ experienced epitaxial growth, with larger columnar dendrites aligned nearly perpendicular to the weld center line, as shown in
Figure 2b. In contrast, coarse dendrites were noticed, as shown in
Figure 2c.
Figure 2b illustrates the interface region, highlighting the narrow HAZ within the dashed yellow lines because of a lower heat input during laser-welding. The grains in the HAZ underwent coarsening due to an insufficient heat input, as noticed in
Figure 2b.
Figure 2c presents the microstructure of the NiTi WM, mainly with columnar dendrites. The WM retained the B2 austenitic structure and formed Ti
2Ni and Ni
4Ti
3 precipitates [
30]. Identical microstructural characteristics were reported in earlier studies [
6,
31].
Considering the microstructural differences across the laser-welded NiTi, an EDS area scan analysis was performed at the BM, interface (IF), and WM regions, as shown in
Figure 3. The IF region was comprised of the BM, HAZ, and WM (refer to
Figure 3b). The concentrations of Ni, Ti, and Cu elements were obtained at different locations (refer to
Table 4). There was not much difference in the concentration of Ni and Ti across the weldment. A minor difference was noticed, confirming the quality of laser-welded NiTi sheets and the formation of precipitates. The obtained EDS results align with Mehrpouya et al.’s previous observation [
6]. As a result of the vaporization phenomenon during laser-welding, the wt% of Ni in the HAZ and WM marginally decreased [
32].
Table 4 also makes it clear that there was very little variation in the weight percentage of Ni and Ti throughout the weld, i.e., lower heat input will have less influence on the weld, resulting in the preservation of functional properties. The difference in the Cu levels at different regions was attributed to the differences in the melting point, solubility, and partition coefficients compared to other elements. During welding, the rapid solidification may have been distributed differently in various regions of the NiTi WM.
Figure 4 shows the SEM image and EDS line scan at the WM, highlighting the concentration of Ni and Ti elements. The area highlighted by the red dotted lines indicates that metastable Ti
2Ni and Ni
4Ti
3 precipitated phases were present in the WM. These phases have been previously reported in earlier studies [
33]. Hence, the decline in Ni and increase in Ti was noticed within the WM region compared to the BM, while the same can be confirmed from the EDS line-mapping spectra. The increase in the Ti was attributed to the recirculation of the molten pool during laser-welding [
24].
Figure 5 shows the XRD spectra of different intermetallic phases formed in the NiTi specimen after laser-welding. Equiatomic NiTi was the main phase, along with the intermetallic precipitates, such as Ti
2Ni (refer to
Figure 5a) and Ni
4Ti
3 (refer to
Figure 5b). The presence of Ti
2Ni affected the mechanical properties. In contrast, the existence of Ni
4Ti
3 affected the phase transformation temperatures and mechanical characteristics. The NiTi WM samples clearly showed two distinct peaks of intermetallic precipitates, including Ti
2Ni (222) and Ni
4Ti
3 (84-2). This observation was similar to the results obtained in the previous study [
24,
34].
The hardness variation across the NiTi weldment is shown by the contour plot along with the microstructure and indentations at different zones in inset images (refer to
Figure 6). The macrograph was matched with the contour map to highlight the hardness variation at the BM, HAZ, and WM. The average hardness in the NiTi BM was 352 ± 5 HV. Because of the microstructural variations throughout the NiTi weldment, the microhardness value progressively decreased from the BM to the WM center through HAZ. The hardness values in the HAZ and WM ranged between 275 and 307 HV and 265 and 287 HV, respectively. The grain coarsening in HAZ and dendrites in the WM reduced the hardness considerably. During solidification, the nucleation and growth of dendrites, the formation of Ti
2Ni and Ni
4Ti
3 precipitates, and the recrystallization of HAZ influenced the hardness in the NiTi joint. Identical trends in hardness measurements have been reported in earlier studies [
16,
35].
Figure 7 represents the engineering stress vs strain curves of BM and WM specimens of NiTi. The average tensile characteristics reported in
Table 5 were determined by testing three samples each from the BM and WM. The graphs show that the WM specimens’ tensile strength (UTS) and percentage of elongation (EL) were much lower than those of the BM specimens. The average UTS and EL of the BM specimen was 1430 ± 34 MPa and 34.50 ± 1%, respectively. In contrast, the WM samples had a UTS and EL of 481 ± 19 MPa and 13.60 ± 0.5%, correspondingly due to the microstructural difference. The microstructural difference and the formation of precipitated phases with Ni and Ti after laser-welding corroborate the considerable decrease in the tensile properties. The BM and WM NiTi specimens ruptured at the sample’s center. Identical trends have been reported in the previous studies focused on NiTi alloys via laser-welding [
16].
The SEM micrographs of the fractured tensile specimens of the NiTi BM and WM are illustrated in
Figure 8a and
Figure 8b, respectively. The fracture mechanism in the BM is ductile in nature with dimples and micro-voids, as the ductility is better. However, the WM specimen underwent a drastic reduction in ductility, and the same can be confirmed by the ruptured surface characteristics (refer to
Figure 8b). The fracture mechanism in the WM was brittle in nature with transgranular cleavage facets. No significant flaws were found on the BM and WM ruptured surfaces.
Electrochemical corrosion tests have been used to evaluate the localized corrosion resistance of the NiTi BM and WM using corrosion potential measurements. The NiTi BM and WM specimens’ open circuit potential (OCP) in an SBF environment is plotted against time in
Figure 9. Approximately 55 mV was the OCP difference between the NiTi BM and the NiTi WM. This suggests that the formation of a protective oxide layer on the NiTi BM sample was more stable than that of the oxide layer that developed on the WM. In addition, there were no spikes in the plot, highlighting the absence of metastable pits.
Figure 10 shows the potentiodynamic polarization (PDP) curves, and
Table 6 describes the corrosion potentials of NiTi BM and WM samples in SBF. It indicates the localized passivation behavior of the NiTi BM and WM specimens from the corrosion current density (I
corr) measurements. The sample that exhibited a higher corrosion potential (E
corr) and lower Icorr exhibited superior passivation performance in corrosive environments [
36]. The NiTi BM sample had better corrosion resistance than the WM sample. For both the BM and WM specimens of NiTi, the Tafel plot shifted from the cathodic to the anodic regions, with different E
corr and I
corr values. Evidently, laser-welding reduced the resistance to corrosion in the NiTi alloys, as indicated by a lower E
corr, higher I
corr, lower E
pit, and higher corrosion rate, as reported in
Table 4. The pitting potential (E
pit) of the NiTi BM and WM were 281.43 mV and 330.05 mV, respectively. According to the corrosion principle, the specimen with a positive E
pit value has a high level of resistance to localized corrosion and can withstand the breakdown of passive film [
37]. The lower E
pit value of NiTi BM highlights the superior resistance to localized corrosion compared to the WM. The average corrosion rate of the NiTi BM and WM specimens in mils penetration per year (mpy) was 0.048 ± 0.0018 mpy and 0.41 ± 0.019 mpy, respectively. The corrosion rate was higher in the WM than in the BM specimen, and it mainly progressed in the Ti-enriched region. Fontana’s classifications indicate that a corrosion rate of less than 1 mpy is outstanding for the most widely used stainless steels and nickel-based superalloys, hence confirming the superior corrosion performance of the NiTi BM and WM [
38].
Using Nyquist plots (refer to
Figure 11) obtained from the electrochemical impedance spectroscopy (EIS) analysis, the electrochemical behaviors of the NiTi BM and WM were examined. The diameter of the semicircles of the NiTi BM sample was significantly bigger than that of the WM, as can be seen from the Nyquist spectra displayed in
Figure 11. The inset graphic in
Figure 11 illustrates the equivalent circuit utilized for fitting the Nyquist curves. This indicates that NiTi BM had superior corrosion resistance to the NiTi WM sample. The capacitive spectra of NiTi WM confirm the lower corrosion resistance as a result of welding. The Nyquist spectra’s semi-circular capacitive arc indicates the passive film formation on the specimen surface, confirming the capacitance characteristics [
29,
39]. The corroded electrode’s maximum charge transfer resistance is highlighted by the larger capacitance radius, which also exhibited superior corrosion resistance [
40]. The Nyquist spectra, OCP, and PDP curves of the NiTi specimens showed a good correlation.
Figure 12 and
Figure 13 show the SEM images of the pits and their corresponding EDS elemental maps around the pits after corrosion tests. Pits were noticed in both the NiTi BM and WM specimens. The size of the pits was less in the BM compared to the WM sample, and this is in good agreement with the corrosion rate. EDS elemental maps confirmed the depletion of Ni and Ti from the specimen surface exposed to an SBF environment, and the formation of oxides can be confirmed. Evidently, the release of Ni was confirmed when the metals were exposed to SBF environments for a prolonged duration. The formation of TiO
2 oxides in the BM provides higher corrosion resistance, while the formation of precipitates, along with a coarse microstructure, affects the corrosion resistance of the WM. The main elements present in the corroded surface were nickel (Ni), titanium (Ti), oxygen (O), sodium (Na), and chlorine (Cl). The elements other than Ni and Ti on the specimen surface were corroborated by the interaction between the SBF and NiTi samples, resulting in the formation of oxides. The presence of O was noticed in the Ni- and Ti-depleted regions around the pits. A severely corroded region could be observed in the NiTi WM, as shown in
Figure 12. From a previous study, it was noticed that the passive film formed on the NiTi BM and WM specimens was mainly comprised of highly stable TiO
2 and acted as a barrier against the oxidation of Ni [
41].