Abstract

The effect of rare earth element Nd addition on the microstructure and martensitic transformation behavior of Ni50Ti50−xNdx (, 1, 3, 7, 20) shape memory alloy was investigated experimentally. The results showed that the microstructure of Ni-Ti-Nd ternary alloy consists of the NiNd phase and the NiTi matrix. One-step martensitic transformation was observed in all alloys. The martensitic transformation start temperature increased gradually with increasing Nd content for Ni-Ti-Nd alloys.

1. Introduction

Ni-Ti based shape memory alloys (SMAs) have unique shape memory effects and super-elasticity behavior and have been used in various fields, particularly in engineering and medical applications [1]. Current research interest on SMAs mainly lies in controlling the martensitic transformation temperature and improving the shape memory effect for their applications. The effects of martensitic transformation, super-elasticity, and shape memory have been widely studied by adding transitional elements to Ni-Ti binary alloys. These elements include Fe [2], Hf [3], and Pd [4]. It is found that most alloying elements such as Fe lower martensitic transformation temperature, but there are only a few, such as Hf and Pd, which increase martensitic transformation temperature [5].

Moreover, the microstructure and martensitic transformation temperature of the rare earth (RE) of Ce [6], Gd [7], Dy [8], and La [9] in addition to Ni-Ti binary alloys have also been studied using scanning electron microscopy (SEM), energy dispersive spectrometry (EDS), X-ray diffraction (XRD), and differential scanning calorimetry (DSC). The addition of these REs to Ni-Ti binary alloys was found to increase the martensitic transformation temperature and change the phase transformation sequence.

RE element Nd is also a widely used element, particularly in magnetic materials. However, only few studies have been conducted on Nd addition to shape memory alloy. The only study found in the literature is that on Nd addition to Ni-Ti-Fe alloy, but the Nd fraction is less than 1 at.% [10]. The effect of Nd addition to Ni-Ti binary alloy on microstructure and martensite transformation temperature remains unclear. In this paper, Nd content with atomic fractions of 1%, 3%, 7%, and 20% was added to Ni-Ti binary alloys, and the microstructure and martensitic transformation were studied experimentally.

2. Material and Methods

The Ni50Ti50−xNdx alloys were prepared by melting each 10 g of raw materials with different nominal compositions (99.9 mass% sponge Ti, 99.7 mass% electrolytic Ni, and 99.95 mass% Nd) in a nonconsumable arc-melting furnace using a water-cooled copper crucible. The alloys are denoted by Nd0, Nd1, Nd3, Nd7, and Nd20 to refer to Ni50Ti50, Ni50Ti49Nd1, Ni50Ti47Nd3, Ni50Ti43Nd7, and Ni50Ti30Nd20 alloys, respectively. Arc-melting was repeated four times to ensure the uniformity of composition. The specimens are spark-cut from the ingots and solution-treated at 850°C for an hour in a quartz tube furnace. Subsequently the specimens were quenched using water. Thereafter, the specimens are mechanically and lightly polished to obtain a plain surface.

The phase transformation temperatures of Ni50Ti50−xNdx alloys were determined by DSC using a TA Q2000 calorimeter. The temperature range of heating and cooling was from −30°C to 140°C, and the scanning rate of heating and cooling was 10°C/min. SEM observations were conducted using a FEI Quanta 650 FEG equipped with EDS analysis systems made by Oxford. An XRD experiment was conducted using a D/MAX-2500PC X-ray diffractometer. The transmission electron microscope (TEM) sample was prepared using the standard technique involving mechanical grinding followed by Ar+ ion milling. TEM experiment was conducted on a JEM-2010 TEM at 200 kV. Images were recorded on a Gatan 1k ×1k slow scan CCD camera and processed using the Gatan DigitalMicrograph software.

3. Results and Discussion

3.1. Microstructure of Ni50Ti50−xNdx Alloy

Figure 1(a) depicts the XRD curves of Ni50Ti50−xNdx alloys at room temperature. The diffraction peaks are identified to be from NiTi B19′ martensite phase, NiTi B2 austenite phase, NiTi2 phase, and NiNd alloy after comparing with JCPDF cards (numbers 65-0145, 65-4572, 72-0442, and 19-0818). The detailed crystal plane indices are marked in Figure 1(b) for Nd0 and Figure 1(c) for Nd3, but the relative intensities of each XRD curve are quite different because of the differences in martensite phase fraction and austenite phase fraction. In this paper, the letter M denotes the NiTi B19′ martensite phase and the letter A denotes the NiTi B2 austenite phase. This perspective will be confirmed in the following DSC analysis. Figure 1(d) depicts a comparison of diffraction peaks of Ni50Ti50−xNdx   alloys. The peak is observed to move toward the left for Ni50Ti50−xNdx alloys; that is, the diffraction angle decreases with increasing Nd fraction, which indicates that the interplanar spacing of (111) of the martensite expands with Nd addition from 1 at.% to 20 at.%. The lattice parameters of alloys can be also calculated by peaks’ position in XRD curves and shown in Table 1. It is shown clearly that the cell volume expands for either martensite or austenite with Nd addition to Ni-Ti binary alloy from 0 at.% to 20 at.%. The observation can also be confirmed in the following composition analysis.

3.2. Morphologies and Compositions of Ni50Ti50−xNdx Alloys

Figure 2 depicts the backscattering SEM images of Ni50Ti50−xNdx () alloys. For Nd0 alloy, there are two different morphologies, namely, black phase and matrix, which can be identified in the SEM image (Figure 2(a)). The black phase is in irregular shape and distributed randomly in the matrix. For Nd1, Nd3, and Nd7, two different morphologies, namely, white phase and matrix, can be identified in the SEM images (Figures 2(b)2(d)). Some white particles that are nearly round-shape and up to 4 μm, 6 μm and 11 μm in diameter, respectively, with white curving areas can be found to be distributed randomly in the matrix. When the Nd content is 20 at.%, the white phase mainly takes on very irregular shape and becomes larger and interconnected, while some white particles that are nearly round-shaped and up to 14 μm are also found to be distributed in the matrix in Figure 2(e). The size of the white particles evidently increases, whereas the white phase overspreads steadily on the matrix with increasing Nd fraction.

To identify the boundaries between the white phase and the matrix, TEM observation was conducted for Nd1 alloy and the TEM bright field images of a small nearly round-shaped particle with a diameter of 1.7 μm and a large nearly round-shaped particle with a diameter of 3.6 μm are shown in Figure 3. The boundaries between these two particles and the matrix look very bright and can be identified clearly at such a very low magnification, which means that the boundaries are incoherent. Furthermore, Xu et al. proposed that the binding force of RE elements and Ni is much bigger than that of Ni and Ti [11]. So, NiNd phase is formed first and then NiTi phase crystallization occurs during the Ni-Ti-Nd alloys preparation. Thus NiNd phase and NiTi matrix have no orientation relationship [5].

To identify the phase structure, EDS analysis was conducted in SEM. The compositions of Ni-Ti-Nd alloys are shown in Table 2. The Ti : Ni ratio in the matrix of all Ni-Ti-Nd alloys is measured to be near 1. Thus, the matrix can be concluded to be NiTi phase with a small amount of Nd solute. The Ti : Ni ratio in the black phase of Nd0 alloy is measured to be near 2 : 1. By XRD analysis, there is a NiTi2 phase in Nd0. Thus, the black phase can be concluded to be NiTi2 phase. According to the 773 K isothermal section of the ternary alloy phase diagram of the Ni-Ti-Nd, no intermetallic compounds can be found in the Ti-Nd binary system. However, Ni-Nd binary alloy phase diagram shows seven kinds of intermetallic compounds defined as NdNi5, Nd2Ni7, NdNi3, NdNi2, NdNi, Nd7Ni13, and Nd3Ni [12]. The EDS results show that the Ni : Nd ratio in the white phase is near 1 and can be regarded as the NiNd intermetallic compound with a small amount of Ti solute. The amount and size of the NiNd phase increase with increasing Nd fraction (Figure 2).

Furthermore, Nd content has also been found in the matrix of Nd3 Nd7, and Nd20. The Nd atomic radius (0.206 nm) is known to be larger than Ti atomic radius (0.176 nm) by 17% and Ni atomic radius (0.149 nm) by 38% [13]. The Nd atom occupies the position of Ni or Ti, consequentially resulting in an expansion of the NiTi matrix lattice [7], which is consistent with the previous XRD analysis.

3.3. Phase Transformation of Ni50Ti50−xNdx Alloys

Figure 4(a) depicts the DSC curves of the Ni50Ti50−xNdx   alloys. Each DSC curve of Nd0, Nd1, Nd3, Nd7, and Nd20 shows only one peak during the heating and cooling process, which indicates a one-step B2B19′ phase transformation. Figure 4(b) shows the effect of Nd concentration on martensitic transformation start temperature . For Nd0 alloy, the is measured to be 77.44°C. It is well known that quenched Ni-Ti binary alloys show one-step B2B19′ transformation and the transformation temperatures are strongly dependent on Ni concentration [5, 7]. 0.1 at.% increase in Ni concentration can lower the of Ni-Ti binary alloys by more than 10°C. For example, Liu et al. measured the to be about −50°C for Ni50.7Ti49.3 alloy after annealing at 900°C for 60 min [7]. Tabish et al. measured the to be −22.12°C for Ni50Ti50 alloy after annealing at 1000°C for 120 min [14]. Wasilewski et al. measured the to be 65°C for Ni49.8Ti50.2 alloy [15]. In this work, the composition of the matrix in Nd0 is measured to be Ni49.36Ti50.64, which is Ti-rich. So, a high of Ti-Ni binary alloy Nd0 is reasonable. Again, the martensite transformations finish temperature in Nd0 alloy is higher than room temperature of 20°C. Thus, the martensite transformations have finished at room temperature and the Nd0 alloy should be in total martensite phase, which is in agreement with the XRD results.

Meanwhile, as shown in Figure 4, the increases with increasing Nd fraction from 1 at.% to 20 at.%. And all in four DSC curves of Nd addition alloys are lower than room temperature. Thus, martensite transformation cannot finish fully at room temperature, which indicates that both the austenite phase and the martensite phase exist in the Ni-Ti-Nd alloy. For Nd1 alloy, the is higher than the others, after which the martensite transformation has nearly finished. Thus, a higher martensite fraction than austenite fraction is present. Consequently, the XRD curve of Nd1 shows that the intensities of the martensite diffraction peaks are significantly higher than those of austenite in Figure 1(a). With increasing Nd fraction, the , the martensite fraction in the alloy, and the relative XRD intensities of the martensite diffraction peaks also decreased.

4. Conclusions

In summary, the effect of RE element Nd addition on the microstructure and martensitic transformation behavior was investigated by XRD, SEM, TEM, and DSC. The microstructure of the Ni50Ti50−xNdx alloys consists of NiNd alloy with a small amount of Ti solute and NiTi matrix with a small amount of Nd solute. The lattice of NiTi matrix is expanded by Nd addition. The Ni-Ti-Nd alloy has a one-step martensitic transformation. Increasing the Nd fraction, the martensitic transformation start temperature increases gradually.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grant no. 51261017), the Program for New Century Excellent Talents in University (Grant no. NCET-10-0909), and the Specialized Research Fund for the Doctoral Program of Higher Education (Grant no. 20101514110002).