Martensitic Transformation and Magnetic Properties of Ni-Co-Mn-In-Gd Ferromagnetic Shape Memory Alloys

Abstract

The effect of rare earth Gd on the microstructure, martensitic transformation and magnetic properties of Ni₄₅Co₅Mn₃₅In_15−xGd_x (0 ≤ x ≤ 1.6) alloys is investigated. The addition of Gd results in a change in the microstructure. With the increase of Gd content, the grain size is clearly reduced and the volume fraction of the second phase increases gradually. And then the second phase along the grain boundaries grows and connects to each other. Some disperse in the grains. One-step thermoelastic martensitic transformation is observed in Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys. When the Gd content reaches 0.8 (at%), the martensite transformation temperature decreases first, and then rises with the increase of Gd content. The near-room-temperature magnetocaloric effect (MCE) in polycrystalline Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloys is observed. The magnetic-field-induced inverse phase transformation from martensite to austenite phase is confirmed. The inverse magnetic entropy change (ΔS_M) reaches 17.78 J/kg K for Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy at 277 K and its R_c is about 356 J Kg⁻¹ at an applied field of 5 T.

1. Introduction

The Ni-Mn based Heusler alloys with martensitic transition have attracted a lot of interest in recent years due to potential applications^1–6). Ni-Mn-In Heusler alloys exhibit unique shape recovery behaviors due to the magnetic-field-induced phase transformation and perfect shape memory effect (SME) of about 7% with almost 100% of shape recovery rate²⁾. Magnetocaloric effect (MCE) is drawing great attention because of its potential application in magnetic refrigeration. Compared with a traditional gas refrigeration method, the magnetic refrigeration shows advantages such as compactness, high efficiency and environmental friendship⁷⁾. To achieve low cost production and high MCE at room temperature, Ni-Mn based Heusler alloys have also been attracting considerable attention⁸⁾, including conventional MCE in Ni-Mn-Ga⁹⁾ and inverse MCE in Ni-Mn-In, Ni-Mn-Sn, and Ni-Co-Mn-In¹⁰⁾. Gd has the highest Curie temperature in the Rare earth elements. The characteristic of Gd with 4f electronic layer may improve MCE of Ni-Co-Mn-In-Gd alloys. However, little information about heavy rare earth Gd-doped Ni-Co-Mn-In alloys is available up to now. Therefore, it is of much importance for further investigation on such compositional polycrystalline Ni-Co-Mn-In-Gd alloys. The aim of the present paper is to investigate the influence of Gd content on the structural and magnetic transition temperatures in Ni₄₅Co₅Mn₃₅In_15−xGd_x (0 ≤ x ≤ 1.6) alloys.

2. Material and Method

Ni₄₅Co₅Mn₃₅In_15−xGd_x (0 ≤ x ≤ 1.6) polycrystalline were prepared by arc melting elemental mixture of Ni (99.95%), Co (99.95%), Mn (99.92%), In (99.95%), and Gd (99.95%) under argon atmosphere, then cast into rods with 10 mm in diameter. The rods were homogenized at 1173 K for 12 h in a sealed quartz tube, and then quenched in cold water. The microstructures of alloys were examined use a Hitachi S-4800 scanning electron microscope (SEM) equipped with an energy dispersive X-ray spectroscopy (EDS) analysis system. The phase transformation temperatures of this alloy were measured by differential scanning calorimetry (DSC, PerkinElmer). Magnetic measurements were taken by physical properties measurement system (PPMS, Quantum Design) in an applied field up to 5 T.

3. Results and Discussion

Figure 1 shows the backscattered electron images of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys (uncorroded). Obviously, the addition of Gd significantly changes the microstructure of Ni₄₅Co₅Mn₃₅In₁₅. As shown in Fig. 1, the Ni₄₅Co₅Mn₃₅In₁₅ alloy is monophase, while the microstructure of the alloys containing Gd consists of the matrix (black area) and the second phase (white area). When the content of Gd is 0.8 at%, a small amount of the second phase disperses in the grains and some distribute along the grain boundaries. When x = 1, more the second phase disperses in the grains and gathers along the grain boundaries compared with Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8. As the content of Gd increases, the second phase interconnects gradually and tends to segregate at the grain boundaries. It also can be seen that the volume fraction of the second phase rises gradually with the increase of Gd content. When x = 1.6, the second phase along the grain boundaries is fully interconnected and the grain size is the smallest compared to others, which indicates that the grain reduce with the increase of Gd content. The grain size of Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy is almost twice bigger than Ni₄₅Co₅Mn₃₅In_13.4Gd_1.6. Compared with the microstructure in Ni-Mn-Ga alloys including rare earth element^11,12), the Ni₄₅Co₅Mn₃₅In_15−xGd_x (0 ≤ x ≤ 1.6) alloys indicate a similar change of the microstruture except for the existence of more granular second phase inside the grains.

Fig. 1

Blackscattering electron images of Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys: (a) x = 0; (b) x = 0.8; (c) x = 1; (d) x = 1.2; (e) x = 1.4; (f) x = 1.6.

The Table 1 shows the EDS results of Ni₄₅Co₅Mn₃₅In_15−xGd_x. It can be see that there is no Gd detected in the matrix. This means that the solubility of Gd in the matrix is very low. In the matrix, with the increase of Gd, the content of Ni and Co keep almost unchanged and Mn content increases, whereas In content decreases. Compared with the composition of the matrix, the second phase has a larger concentration of Gd and In, and a smaller concentration of Mn and Ni, while the Co content remains almost unchanged. As shown in the Table 1, Gd content is almost constant in the second phase, which indicates insolvable Gd element form the Gd-rich phase.

Table 1 The EDS results of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1.0, 1.2, 1.4, 1.6) alloys.

Phase	Rare earth content x	Ni (at%)	Co (at%)	Mn (at%)	In (at%)	Gd (at%)
Matrix	0	45.1	5.1	35.5	14.3	-
	0.8	45.8	5.0	35.9	13.3	-
	1	44.8	5.2	37.8	12.2	-
	1.2	46.6	4.8	36.7	11.8	-
	1.4	47.4	5.1	37.2	10.3	-
	1.6	45.6	5.8	38.8	9.8	-
The second phase	0.8	33.2	3.2	12.4	38.3	12.9
	1	40.3	4.9	11.9	28.3	14.6
	1.2	36.4	4.4	14.1	31.8	13.3
	1.4	39.4	4.1	12.2	32.6	14.7
	1.6	36.2	3.9	12.8	33.7	13.7

X-ray diffraction patterns of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys at room temperature are shown in the revised manuscript. Figure 2 shows the X-ray diffraction patterns of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys at room temperature. For the Ni₄₅Co₅Mn₃₅In₁₅ and Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloys, the X-ray diffraction patterns are all indexed to coexistence of 10M modulated martensite structure and L2₁ cubic austenite structure. It is consistent with the DSC results. Whereas, when the content of Gd is between 1 at% and 1.6 at%, 14M structure is observed and the main reflection is indexed on a monoclinic structure. The lattice parameters and crystal phase for Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys are summarized in Table 2. It is worth noting that when the content of Gd is 1 at% or above, in addition to the diffraction peaks of martensite phase, some additional peaks appears, as indicated by the arrows in Fig. 2. The intensity of the additional peaks increases slightly with increasing Gd content, implying that a new phase a formed. This suggests that the new phase may be the Gd-rich phase mentioned above. The crystal structure of the Gd-rich second phase needs to be investigated in further by TEM and XRD at different temperature.

Fig. 2

X-ray diffraction patterns of Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys at room temperature: (a) x = 0; (b) x = 0.8; (c) x = 1; (d) x = 1.2; (e) x = 1.4; (f) x = 1.6.

Table 2 The lattice parameters and crystal phases Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys.

Composition	a(Å)	b(Å)	c(Å)	β(°)	Crystal phase
Ni45Co5Mn35In15	4.44	5.96	20.49	90.25	10M + A
Ni45Co5Mn35In14.2Gd0.8	4.45	6.02	20.40	91.21	10M + A
Ni45Co5Mn35In14Gd1	4.82	5.70	30.22	89.60	14M
Ni45Co5Mn35In13.8Gd1.2	4.31	5.96	30.16	93.90	14M
Ni45Co5Mn35In13.6Gd1.4	4.33	5.97	30.22	88.53	14M
Ni45Co5Mn35In13.4Gd1.6	4.26	6.74	30.06	90.72	14M

The heating and cooling DSC curves for each alloy near the martensitic transformation temperature are presented in Fig. 3. It can be seen that an exothermic peak and an endothermic peak occur upon cooling and heating for the Ni-Co-Mn-In-Gd alloys. That is the typical characteristic of one-step thermoelastic martenstic transformation. Figure 4 shows the influence of Gd content on the martensite transformation temperature of Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys. When x = 0, the martensitic transformation start temperature (M_s) is determined to be 286 K. When the Gd content reaches 0.8 (at%), the martensite transformation temperature decreases first, and then rises with the increase of Gd content. When the content of Gd is 1.6 at%, the martensitic transformation start temperature is up to 394 K (about 108 K of Ni₄₅Co₅Mn₃₅In_13.4Gd_1.6) and still has a tendency to increase. Based on our previous results, the martensitic transformation temperatures were increased by the substitution of Gd on Ga in Ni-Mn-Ga alloys. However, just when the content of Gd is up to 2 at%, the M_s of Ni₅₀Mn₂₉Ga₁₉Gd₂ alloy is raised to 390 K, which is about 64 K of the ternary Ni₅₀Mn₂₉Ga₂₁ alloy¹²⁾. It indicates that the effect of increasing martensitic transformation temperature by the substitution of Gd for In in Ni-Co-Mn-In alloys is more obvious compared with the one of Ni-Mn-Ga-Gd alloy. The change of matrix composition is the main reason for the increase of martensite transition temperature. The EDS results from the Table 1 reveal that the content of Ni and Co in the matrix keeps almost unchanged and the content of Mn increases from 35.5 at% for Ni₄₅Co₅Mn₃₅In₁₅ alloy to 38.8 at% for Ni₄₅Co₅Mn₃₅In_13.4Gd_1.6 alloy. At the same time, the In content in the matrix decreases from 14.3 at% for Ni₄₅Co₅Mn₃₅In₁₅ alloy to 9.8 at% for Ni₄₅Co₅Mn₃₅In_13.4Gd_1.6 alloy. The phase transformation temperatures rise with decrease of content of In in the matrix. This implies that the decrease of In content may be responsible for the increase in the martensitic transformation temperature. This agrees with the results obtained by Y. Feng et al. who showed that the martensitic transformation temperatures increased monotonically with the decrease of In content in Ni₅₀Mn₃₄In₁₄Fe₂ alloys¹³⁾. The reason for the decrease of martensitic transformation temperatures for the Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloys needs to be further investigated in the future.

Fig. 3

DSC curves of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys.

Fig. 4

The effect of Gd content on phase transformation temperatures of Ni₄₅Co₅Mn₃₅In_15−xGd_x (x = 0, 0.8, 1, 1.2, 1.4, 1.6) alloys.

Figure 5 shows the temperature dependence of the magnetization M-T for the Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy in the fields of 0.1 T and 5 T. At a low magnetic field (0.1 T), as the temperature decreases, a sharp increase in magnetization is observed, corresponding to the paramagnetic magnetic transition of the austenite Curie temperature. The T_c, defined as the temperature where the dM/dT is the minimum, as shown in the inset of Fig. 5(a), is 375 K for Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy. A further reduction in temperature results in a sudden drop in magnetization, which corresponds to martensitic transformation. During the heating process, the reverse martensitic transformation from martensite phase to austenite phase is confirmed by the magnetization jump. From Fig. 5(a), upon heating, reverse martensitic temperature starts at A_s = 257 K and ends at A_f = 296 K. Upon cooling, martensitic temperature starts at M_s = 276 K and ends at M_f = 240 K. From Fig. 5(b), the magnetic field of 5 T is applied at 150 K, and the sample is heated to 380 K, and then cooled to 150 K. The curves measured at 5 T external magnetic field are slightly different from the curves at 0.1 T magnetic field, as shown in Fig. 5(b). It can be seen that all the transition temperatures are reduced in the 5 T magnetic field compared to 0.1 T. That indicates the magnetic-field-induced austenite transformation. The transformation hysteresis which is defined as A_f - M_s are 20 K and 23 K, respectively, at the applied fields of 0.1 T and 5 T. The difference of magnetization between martensite and austenite is 0.8 emu/g at the field of 0.1 T, but abruptly increases to about 87 emu/g with increasing magnetic field to 5 T.

Fig. 5

Temperature dependence of magnetization in the Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy under magnetic fields 0.1 T and 5 T.

Isothermal magnetization curves M-H for Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy are measured by 244 K to 292 K, which is prepared by Fig. 6. At each temperature, the sample is magnetized from 0 to 5 T, and then demagnetized from 5 to 0 T. Temperature is increased by 3 K after the cycle is completed. The curves which are measured from 244 K to 289 K reveal that the magnetization of martensitic phase is lower than that of parent phase. The metamagnetic transition behavior which is observed in the sample indicates the presence of magnetic field induces reverse martensitic transition (martensitic to austenite). At the temperature of 292 K, the magnetization is saturated at a very low magnetic field, which indicates that the alloy is completely austenitized, where the M-H curves are reversible.

Fig. 6

Magnetization curves of the Ni₄₅Co₅Mn₃₅In_14.8Gd_0.8 alloy at different temperatures.

The magnetic entropy change (ΔS_M) is calculated based on the isothermal magnetization data by using the Maxwell relation:   

\[{\rm \Delta S_{M}} = \int_{H_I}^{H_F} \left(\frac{\partial M(T,H)}{\partial T}\right)_H \ dH\]

H_I and H_F are the initial and final magnetic field, respectively.

Figure 7 shows the magnetic entropy change as a function of temperature for Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy for various changes in the applied magnetic fields. At the field of 5 T, the peak of ΔS_M appeals is 17.78 J/kg K at 277 K. The magnetic entropy may be weakened by the Gd-rich phase. This agrees with the results obtained by J. Liu et al. who showed that the formation of the ferromagnetic γ phase smears the martensitic transformation and dilutes the MCE¹⁴⁾.

Fig. 7

ΔS_M as a function of temperature of the Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy for various changes in applied magnetic field.

The ΔS_M position shows a field dependence behavior, i.e. the peak ΔS_M shifts to a lower temperature with the increase of magnetic field. One possible reason for it may be related to the intersection of the M-H curves around A_s. As shown in Fig. 6, the M-H curves of 286 K and 289 K intersect at 3.5 T, below which the increase of magnetic field leads to a positive contribution to ΔS_M. As a result, the maximum ΔS_M for a field change at the magnetic field of 5 T is only achieved at the lower temperature, i.e. 277 K, where the adjacent M-H curves of 274 K and 280 K are not intersected. This agrees with the results obtained by J. Liu et al. who showed that the intersection of the M-H curves around A_s may be responsible for the shifts that peak ΔS_M is transferred to a lower temperature as the magnetic field changes¹⁴⁾.

The isothermal M-H curves are recorded up to 5 T, as shown in Fig. 8. For the sample at the temperature of 223 K (far below A_s), the magnetic field up to 5 T almost cannot drive the phase transformation, so the field cycling has no influence on the M-H curve. The M-H curves at the temperature of 268 K (9 K below A_s), show that the first cycle can't coincide with the second one. This indicates that a large part of the field-induced austenite is trapped after the first cycle of the sample, which is consistent with the large thermal hysteresis in Fig. 8. It indicates that the magnetic field drives the inverse martensitic transformation. The M-H curves at 273 K (4 K below A_s) are similar with the M-H curves at 268 K. At 323 K (46 K above A_s), the magnetization is saturated at a very low field, which indicates that the alloy is fully austenitized. This result well coincide with the martensite transformation temperature measured in M-T curves.

Fig. 8

Several cycles of isothermal M-H curves of Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy at different temperatures.

The useful parameter to assess the MCE is refrigerant capacity (R_c), which is defined as:   

\[\rm R_{c} = {\rm \Delta S_{max}}{\rm \delta T_{FWHM}}\]

Here, δT_FWHM is the full width at half maximum of the ΔS_M(T) curve.

Although ΔS_max value of Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy is relatively small, ΔS_M(T) curves are distributed over a wide temperature. R_c is about 356 J Kg⁻¹ at the field of 5 T. This value is bigger than the R_c of 334 J Kg⁻¹ of Ni₄₂Co₈Mn₃₈In₁₂ under 5 T¹⁵⁾. The near-room-temperature MCE with a higher R_c reveals the material system may be used in refrigeration application.

4. Conclusion

The effect of Gd addition on the microstructure, martensitic transformation and magnetic properties of Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys was studied. The results show that Gd doping has a clear refining effect on grain, and it forms a Gd-rich phase which contain Ni, Co, Mn, In and Gd. With the increase of Gd content, Gd-rich phase gradually becomes larger, and the trend is mainly distributed on grain boundary, some spread in the grains.Ni₄₅Co₅Mn₃₅In_15−xGd_x alloys has a one-step thermoelastic martensitic transformation. The martensite transformation temperature decreases first, and then increases with the increase of Gd content. The martensitic transformation start temperature (M_S) of Ni₄₅Co₅Mn₃₅In_13.4Gd_1.6 alloy increases about 108 K compared with Ni₄₅Co₅Mn₃₅In₁₅ alloy. For Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8alloy at 5 T, the inverse magnetic entropy change (ΔS_M) reaches 17.78 J/kg K at 277 K and its R_c is about 356 J Kg⁻¹. The larger magnetic entropy change near room temperature in the Ni₄₅Co₅Mn₃₅In_14.2Gd_0.8 alloy is a promising working material for magnetic refrigeration.

Acknowledgements

This work is supported by National Natural Science Foundation of China (No. 51401122) and (No. 51671126).

REFERENCES

• 1)  L. Mañosa, D. Gonzalezalonso, A. Planes, E. Bonnot, M. Barrio, J.L. Tamarit, S. Aksoy and M. Acet: Nat. Mater. 9 (2010) 478–481. 10.1038/nmat2731
• 2)  R. Kainuma, Y. Imano, W. Ito, Y. Sutou, H. Morito, S. Okamoto, O. Kitakami, K. Oikawa, A. Fujita, T. Kanomata and K. Ishida: Nature 439 (2006) 957–960. 10.1038/nature04493
• 3)  H.E. Karaca, I. Karaman, B. Basaran, Y. Ren, Y.I. Chumlyakov and H.J. Maier: Adv. Funct. Mater. 19 (2009) 983–998. 10.1002/adfm.200801322
• 4)  T. Krenke, E. Duman, M. Acet, E.F. Wassermann, X. Moya, L. Manosa and A. Planes: Nat. Mater. 4 (2005) 450–454. 10.1038/nmat1395
• 5)  V.K. Sharma, M.K. Chattopadhyay, K.H.B. Shaeb, A. Chouhan and S.B. Roy: Appl. Phys. Lett. 89 (2006) 222509. 10.1063/1.2399365
• 6)  J. Liu, T. Gottschall, K.P. Skokov, J.D. Moore and O. Gutfleisch: Nat. Mater. 11 (2012) 620–626. 10.1038/nmat3334
• 7)  M.H. Phan and S.C. Yu: J. Magn. Magn. Mater. 308 (2007) 325–340. 10.1016/j.jmmm.2006.07.025
• 8)  L. Pareti, M. Solzi, F. Albertini and A. Paoluzi: Eur. Phys. J. B 32 (2003) 303–307. 10.1140/epjb/e2003-00102-y
• 9)  B. Hernando, J.L.S. Llamazares, J.D. Santos, V.M. Prida, D. Baldomir, D. Serantes, R. Varga and J. González: Appl. Phys. Lett. 92 (2008) 132507. 10.1063/1.2904625
• 10)  K. Oikawa, W. Ito, Y. Imano, Y. Sutou, R. Kainuma, K. Ishida, S. Okamoto, O. Kitakami and T. Kanomata: Appl. Phys. Lett. 88 (2006) 122507. 10.1063/1.2187414
• 11)  K. Tsuchiya, A. Tsutsumi, H. Ohtsuka and M. Umemoto: Mater. Sci. Eng. 378 (2004) 370–376. 10.1016/j.msea.2003.11.076
• 12)  L. Gao, X. Shen and J. Xu: Mater. Trans. 56 (2015) 1186–1191. 10.2320/matertrans.M2015143
• 13)  Y. Feng, J.H. Sui, H.B. Wang and W. Cai: J. Magn. Magn. Mater. 324 (2012) 1982–1984. 10.1016/j.jmmm.2012.01.041
• 14)  J. Liu, T.G. Woodcock, N. Scheerbaum and O. Gutfleisch: Acta Mater. 57 (2009) 4911–4920. 10.1016/j.actamat.2009.06.054
• 15)  F. Chen, Y.X. Tong, L. Li, L. Sanchez Llamazares, C.F. Sanchez-Valdes and P. Muller: J. Alloy. Compd. 725 (2017) 603–609.