Introduction

Recently, more interest has been generated in the ferromagnetic shape memory alloys (FMSMAs) because of their various applications such as sensors and actuators materials.1 A lot of FMSMAs like NiMnGa2–5 exhibit a magnetically induced shape memory effect and excellent thermal stability.6 However, this system of alloys has some drawbacks, mainly originating from the loss of Mn. It causes severe brittleness in polycrystalline alloys in addition to high segregation in single crystal alloys that limits its applications.7,8 Therefore, several trials have been made to develop new systems such as NiFeGa, CoNiAl, and CoNiGa to avoid the demerits of NiMnGa alloys. Accordingly, in comparison with NiMnGa alloys, the new developed group of the FMSMAs in the Co-Ni-Ga alloy system located in the two-phase B2 + γ (martensite + γ) region is characterized by good ductility due to the γ phase in its microstructure.9–12 The microstructure with a higher volume fraction of γ phase, soft phase, at the expense of martensite, hard phase, has higher ductility.

A lot of studies have reported that heat treatment conditions seriously influence the microstructure, martensitic transformation, and magnetic and mechanical properties of various FMSMAs.13–15 In addition to heat treatment influence, Al addition to the CoNiGa system can alter its volume fracture of martensite and γ phase, martensitic transformation temperature peak (M p), and saturation magnetization (M s).

Additionally, it is worth mentioning in this context that valence electron concentration (e v/a) is the most important factor influencing structure, as well as M p, mechanical and magnetic properties such as M s, remanence magnetization (M r), and coercivity (H c).16,17 Furthermore, (ev/a) indicates the number of all electrons in the alloys per number of atoms. Consequently, the following equation is used to calculate this ratio, e v/a, based on the atomic fractions of the alloying elements in the alloy:16

$$ e_{\text{v}} /a \, = f_{1} e^{1}_{\text{v}} + f_{2} e^{2}_{\text{v}} + f_{3} e^{3}_{\text{v}} $$
(1)

where f is the atomic fraction and e v is the number of valence electrons of the elements in the alloy.18–20 There are two opposite opinions that control the relationship between M p and e v/a. For instance, many researchers have reported that the M p is in direct relation with e v/a.16,21–23 Otherwise, others have indicated that it is an inverse relationship between M p and e v/a.24–26 Moreover, it is found that the magnetic saturation measurements are affected by the e v/a. Indeed, the increase in e v/a causes a decrease in M s.17

Herein, in the present study, the influence of aging conditions and Al doping on the microstructure evolution, martensitic transformation, and mechanical and magnetic properties of Co50Ni23Ga27−X Al X alloys is investigated.

Expermiental Procedures

The dual phase of (β + γ) Co50Ni23Ga27−X Al X alloys (X = 0 and 1 at.%) was prepared from the starting materials of 99.99 wt.% Co, 99.99 wt.% Ni, and 99.99 wt.% Ga by arc-melting pure elements under an argon atmosphere in a water-cooled copper crucible. To lower segregation and ensure homogenization as much as possible, the cast alloys used in this investigation were melted four times. After being solutionized at 1273 K for 24 h for both alloys, under a controlled atmosphere of He, these two specimens were sliced into small-sized parts, which were then aged at 823 K for various periods 3 h, 24 h, and 120 h followed by iced water quenching (WQ). The microstructure was evaluated by an optical microscope (OM) (MEIJI TECHNO) fitted with a digital camera and a scanning electron microscope (SEM—JEOL JSM5410). The equilibrium phase composition was detected by an energy x-ray dispersive spectroscopy (EDS) attached to SEM. The crystalline phases in the different samples were distinguished by using x-ray diffraction (XRD) on the Brucker axis D8 diffractometer using Cu Ka (λ = 1.5406) radiation with a step scanning in the 2θ range of 30° to 100°. The peak of martensitic transformation temperatures (M p) was determined by differential scanning calorimetry (Netzsch Leading Thermal Analysis DSC 204 F1) at heating and cooling rates of 10 K/min. Magnetic properties were measured as well by using a vibrating sample magnetometer (VSM; lakeshore 7400; USA). Hardness values were determined by using a Leco Vickers Hardness Tester LV800AT with 31.25 Kgf.

Results and Discussion

Crystal Structure

To investigate the influence of aging heat treatment conditions on the structure and to identify the different phases found in Co50Ni23Ga27 and Co50Ni23Ga26Al1 alloys, x-ray diffraction measurements were performed at room temperature. There are six XRD patterns, as shown in Fig. 1; each alloy was represented by three XRD patterns according to aging at 823 K for various aging times: 3 h, 24 h, and 120 h. The XRD results reveal the crystal structures of the present phases. These phases were determined to be L10 for martensite phase, A1 for γ phase (disordered fcc), and L12 for γ′ phase (ordered cubic). The main reflection (110) was indexed with the tetragonal structure for the parent phase: martensite.25 Moreover, there are some other peaks such as (022), (220), and (211) that indicate the presence of martensite phase as well. Furthermore, γ phase is represented in Fig. 1 by diffraction planes (002) and (420); however, (210) and (311) planes belong to the γ′ phase.

Fig. 1
figure 1

XRD patterns of the aged Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys

Full size image

Microstructure Evolution

Figure 2 depicts the microstructures, observed by optical microscope, of both Co50Ni23Ga27 and Co50Ni23Ga26Al1 alloys where (a), (b), and (c) represent the microstructures of Co50Ni23Ga27 alloys that aged at 823 K for 3 h, 24 h, and 120 h, respectively, and (d), (e), and (f) illustrate the counterpart microstructures of Co50Ni23Ga26Al1 alloy. The microstructures of both investigated alloys aged with different heat treatment conditions are composed of martensite, γ phase in addition to particles of γ′ phase that precipitated in martensite phase, as evinced in Fig. 3. The γ phase and its volume fraction (V f) in the microstructures of Co50Ni23Ga27 alloy, (a), (b), and (c), are coarser and higher than that in the counterpart microstructures of Co50Ni23Ga26Al1 alloy, (d), (e), and (f). Meanwhile, the Vf of the γ phase in both Co50Ni23Ga27 and Co50Ni23Ga26Al1 alloys decreases with prolonging the aging time. Obviously, the Vf of the γ phase diminishes in both alloys with increasing the aging time from 3 h and 24 to 120 h. In addition, γ′ phase is precipitated in the martensite phase in all microstructures of the aged specimens. These particles of γ′ phase got coarser as the aging time increased, as described in Fig. 3.

Fig. 2
figure 2

Microstructure of Co50Ni23Ga27 alloys aged at 550°C for (a) 3 h, (b) 24 h, and (c) 120 h and Co50Ni23Ga26Al1 alloys aged at 823 K for (d) 3 h, (e) 24 h, and (f) 120 h

Full size image
Fig. 3
figure 3

Parent: martensite phase and γ phase and precipitation of γ′ phase in martensite in the microstructure of Co50Ni23Ga27Al1 alloy aged for 120 h

Full size image

The distribution of the alloying elements, Co, Ni, and Ga, in the martensite and γ phases found in the microstructure of Co50Ni23Ga27 alloy aged at 823 K for 120 h is shown in Fig. 4. Chemical distribution provides a clear image about the heat treatment influence on segregation of alloying elements, which in turn affect the presence and the volume fraction of different phases, such as gamma (γ), gamma prime (γ′), and martensite, found in the investigated microstructures. Moreover, chemical distribution explains the anisotropy of the physical and mechanical properties. The martensite phase has higher content of both Ga and Ni elements; however, the γ phase is rich in Co element. Moreover, the microchemical analysis of martensite and γ phases is recorded in Table I, whereas Table II demonstrates the microchemical analysis of γ′ phase. It is noticeable that (γ′) phase is Co based with higher Al content, as presented in Table II. Additionally, this table contains the percentage of error for the measurements of microchemical analyses of alloying elements: Co, Ni, Ga, and Al.

Fig. 4
figure 4

Line analysis through γ and martensite phases in Co50Ni23Ga27 alloy aged for 120 h

Full size image
Table I Microchemical analyses of martensite phase, electron concentration in matrix (e v/a), and martensite transformation temperature (T m) of the investigated alloys
Full size table
Table II Microchemical Analyses of γ′ Phase
Full size table

In addition, the influence of aging treatment conditions on the martensite transformation temperatures (T m) is investigated. As the aging time increases, the T m for both Co50Ni23Ga27 and Co50Ni23Ga26Al1 alloys increase as deduced in Table I. Meanwhile, the T m for Co50Ni23Ga27 alloys is lower than that of Co50Ni23Ga26Al1 alloys. Thus, Al is increasing the martensitic transformation temperature when it is added on account of Ga.26 Moreover, one can note that T m is very sensitive to the chemical composition of the matrix phase, which is martensite; however, there is a percentage of error in any experimental measurements that should be taken into account; for example, see Table II. Accordingly, the Al percentage increases, while the Co content decreases in the matrix, martensite, with aging time. Al and Ga preferably segregate to martensite, while Co selectively separates into the γ phase.6

It is obvious that T m increases with the increase of Ga at.%;1 however, it is decreased by increasing the Co content6 in the matrix of Co50Ni23Ga27 alloy, as shown in Table I. In addition to the effect of Ga and Co, the increase of Al (at.%) elevates the T m of the Co50Ni23Ga26Al1.25

The chemical composition of the matrix affect on the T m is represented by the valence electron concentration (e v/a). A lot of researchers investigated the relationship between T m and (e v/a). This relationship is a controversial issue, where some researchers have indicated that it is a direct relationship.1,16,21,22 However, others have confirmed that it is an inverse relationship.6,23,24 The obtained results in the present study support the latter point of view.

Mehrdad and Yong27 studied this relation in detail. They categorized alloys into low (e v/a < 5), medium (5 ≤ e v/a ≤ 7.50), and high (e v/a > 7.50) valence electron groups. As shown in Table I, the values of (e v/a) in this study belong to the medium category. In this category, according to Mehrdad and Young, the relationship between T m and (e v/a) is an inverse relation that coincides with the results of this study.

Magnetic Properties

The effect of aging time on the magnetic properties of the CoNiGaAl alloys was performed at room temperature under an applied field of 15 kOe using a vibrating sample magnetometer (VSM). These measurements were carried out at room temperature under an applied field of 15 KOe. Plots of magnetization (M) as a function of applied field (H) for the Co50Ni23Ga27 and Co50Ni23Ga26Al1 alloys aged at 823 K for 3 h, 24 h, and 120 h are shown in Fig. 5.

Fig. 5
figure 5

Effect of aging time on the M-H hysteresis loop of CoNiGaAl alloys

Full size image

The change on saturation magnetization (M s), remanence magnetization (M r), and coercivity (H c) for Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys obtained at room temperature are shown in Fig. 6. It is noticeable that M s for both Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys aged at 823 K for 3 h is found to be 124.15 emu/g and 58.36 emu/g, respectively. By prolonging the aging time to 24 h and further to 120 h, the M s for both aged alloys is found to minify into 65.52 emu/g and 50.61 emu/g and then to 53.07 emu/g and 46.16 emu/g, respectively.

Fig. 6
figure 6

Influence of aging time on the saturation magnetization (M s), remanence magnetization (M r), and coercivity (H c) of CoNiGaAl alloys

Full size image

All the measured magnetic properties for aged Co50Ni23Ga27Al0 alloys are higher than in Co50Ni23Ga26Al1 alloys, as shown in Fig. 6. Additionally, the M s decreased for both aged Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys with aging time; however, both M r and H c are increased for aged Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys as the aging time increases.

The main two phases in the microstructures of the investigated alloys are the martensite and the γ phase. These two phases have different magnetic properties, where the γ phase, a Co-rich phase, has larger magnetization compared to martensite phase.28 Moreover, the Co additions increase the V f of the γ phase and make it more stable as well as increase its saturation magnetization that was confirmed by Liu et al.6 However, Al segregates preferably to the martensite phase and increases its V f and lowering of the Ms of aged alloys.

The M s values of aged Co50Ni23Ga26Al1 alloys are lower than that of Co50Ni23Ga27Al0 alloys according to the diminishing in the V f of the γ phase in the microstructure of Co50Ni23Ga27Al0 alloys than of Co50Ni23Ga26Al1 alloys due to the increase in the Al (at.%).25 Furthermore, the decrease in the saturation magnetization for both aged Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys with aging time originated from the decrease in the V f of the γ phase due to the decrease in the Co (at.%).6

Mechanical Properties

The influence of the prolonging time of the aging process on the hardness of the investigated alloys is elucidated in Fig. 7. The hardness values for the specimens aged at 823 K for 3 h, 24 h, and 120 h of Co50Ni23Ga26Al1 alloy, 270, 307, and 353 HV30, are higher than the corresponding specimens for Co50Ni23Ga27Al0 alloys, 262, 277, and 341 HV30, respectively. Hardness measurements for the both investigated alloys are in a direct relationship with aging time, whereas as the aging time increases, the hardness values increase.

Fig. 7
figure 7

Hardness measurements versus aging time

Full size image

The increase in the hardness measurement for both Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys with aging time could be due to the increase of the V f of martensite at the expense of V f of the γ phase. In the same direction, hardness measurements are enhanced with aging time as the result of the precipitation of the γ′ phase in both aged alloys.

In comparison with the counterpart specimens of Co50Ni23Ga27 alloys, Co50Ni23Ga26Al1 alloys have higher hardness values because of the higher Vf of the martensite phase and a finer structure than in Co50Ni23Ga27 alloys.

Conclusion

The microstructure, martensitic transformation, and magnetic and mechanical properties of Co50Ni23Ga27−X Al X alloys that aged at 823 K with various aging times of 3 h, 24 h, and 120 h were investigated. The obtained results are as follows:

  1. 1.

    The microstructure is composed of Ga-rich martensite, as a parent phase, Co-rich γ phase, as well as precipitation of γ′ phase particles for both Co50Ni23Ga27Al0 and Co50Ni23Ga26Al1 alloys.

  2. 2.

    Vf of the γ phase in the Co50Ni23Ga27Al0 alloy is higher than that in the Co50Ni23Ga26Al1 alloy. Moreover, the V f value of the γ phase is decreased by prolonging the aging time for both alloys.

  3. 3.

    The martensitic transformation temperature T m is elevated as the aging time increases; however, it decreases as the electron concentration (e v/a) increases.

  4. 4.

    The magnetic parameters Ms, Mr, and Hc for the Co50Ni23Ga27Al0 alloy are higher than those for the Co50Ni23Ga26Al1 alloy in all aging conditions. Saturation magnetization (M s) values decrease, whereas remanence magnetization (M r) and coercivity (H c) values increase as the aging time increases.

  5. 5.

    Hardness measurements for both alloys are increased with aging time. In comparison with the Co50Ni23Ga26Al1 alloy, the Co50Ni23Ga27 alloy has lower hardness values.