Influence of Ni/Mn ratio on magnetostructural transformation and magnetocaloric effect in Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) ferromagnetic shape memory alloys
Shah Ishfaq Ahmad1, Hassan Najam ul1, Rauf Abdur2, Liu Jun1, Gong Yuanyuan1, Xu Guizhou1, Xu Feng1, †
Jiangsu Key Laboratory of Advanced Micro & Nano Materials and Technology, School of Materials Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
Herbert Gleiter Institute of Nanoscience, Nanjing University of Science and Technology, Nanjing 210094, China

 

† Corresponding author. E-mail: xufeng@njust.edu.cn

Project supported by the National Natural Science Foundation of China (Grant Nos. 51601092, 51571121, and 11604148), the Fundamental Research Funds for the Central Universities, China (Grant Nos. 30916011344 and 30916011345), Jiangsu Natural Science Foundation for Distinguished Young Scholars, China (Grant No. BK20140035), China Postdoctoral Science Foundation (Grant No. 2016M591851), the Natural Science Foundation of Jiangsu Province, China (Grant Nos. BK20160833 and BK20160829), Qing Lan Project of Jiangsu Province, China, Priority Academic Program Development of Jiangsu Higher Education Institutions, China, and NMG–NJUST Joint Scholarship Program for Ishfaq Ahmad Shah (Student ID: 914116020118).

Abstract

An investigation on the magnetostructural transformation and magnetocaloric properties of Ni48–xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) ferromagnetic shape memory alloys is carried out. With the partial replacement of Ni by Mn in the Ni48Co2Mn38Sn12 alloy, the electron concentration decreases. As a result, the martensitic transformation temperature is decreased into the temperature window between the Curie-temperatures of austenite and martensite. Thus, the samples with x = 1.5 and 2.0 exhibit the magnetostructural transformation between the weak-magnetization martensite and ferromagnetic austenite at room temperature. The structural transformation can be induced not only by the temperature, but also by the magnetic field. Accompanied by the magnetic-field-induced magnetostructural transformation, a considerable magnetocaloric effect is observed. With the increase of x, the maximum entropy change decreases, but the effective magnetic cooling capacity increases.

PACS: 75.30.Sg
1. Introduction

Room-temperature magnetic refrigeration is an innovative cooling technology with the advantages of high efficiency, low cost, and environmental friendliness. In this technology, magnetic materials are used as the refrigerants. To maximize the cooling capacity of magnetic refrigeration at room temperature (RT), the magnetic refrigerants should exhibit a large magnetic entropy or temperature change under a magnetic field.[14] During the last decade, the large magnetocaloric effect (MCE) has been successively observed in magnetic alloys that experience a magnetic-field-induced first-order phase transition, including Gd–Si–Ge-based alloys,[5] MnAs1−xSbx,[6] Mn–Fe–P–As,[7] La (Fe1−xSix)13,[8] and especially, Ni–Mn-based magnetic shape memory alloys.[9,10]

Rare-earth-free Ni–Mn-based magnetic shape memory alloys (Ni–Mn–X, X = Al, Ga, Sn, In and Sb) display a structural transformation between magnetic martensite and austenite. Various investigations on the structural, magnetic, and electrical properties of these alloys have been carried out.[1117] It is found that the first-order magnetostructural transformation (MST) from the high-temperature austenite to the low-temperature martensite is accompanied by an obvious magnetization difference (ΔM) in some alloys with specific compositions. Due to the existence of ΔM, the MST can be driven by a magnetic field, and thus a large MCE can be achieved.[1822]

Among Ni–Mn–X alloys, Ni–Mn–Sn attracts considerable attention due to its tunable MST, large MCE, and relatively low cost. From the application point of view, the designed material as a magnetic refrigerant should display the MCE and large effective refrigeration capacity (RCeff) at RT. In this work, we select Ni48Co2Mn38Sn12 as the starting composition, which experiences a structural transformation above RT. By changing the Ni/Mn ratio, the structural transformation temperature (Tt) is decreased to RT. Meanwhile, the structural transformation is tuned into the Curie temperature window, which leads to a large ΔM and be in favor of the magnetic-field-induced structural transformation (MFIST). Based on this improvement, large MCE, reduced hysteresis loss (HL), and improved RCeff are obtained.

2. Experimental details

The polycrystalline Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys were prepared by arc-melting appropriate amounts of high purity elements Ni (99.995 wt.% purity), Co (99.95 wt.% purity), Mn (99.998 wt.% purity), and Sn (99.99 wt.% purity) in argon atmosphere. The samples were re-melted three times for homogeneity. The obtained ingots were annealed at 900 °C in vacuum for 24 h, and then quenched into cold water. The martensitic and reverse transformation behaviors were identified by differential scanning calorimetry (DSC) with the heating/cooling rate of 10 K/min. The crystal structures were investigated by powder x-ray diffraction (XRD) using Cu-Kα radiation at RT. The compositions of the elaborated alloys were confirmed by x-ray energy dispersive spectrometry (EDS). Magnetic measurements were performed on a physical property measurement system (PPMS). Isothermal magnetization (MB) curves were measured using a so-called loop process to avoid the irreversibility caused by the magnetic-field-induced first-order MST.[23] Before each MB measurement, the samples were cooled down to full martensitic state, and then gradually heated to the measurement temperature with a ramp rate of 3 K/min. To ensure the temperature stability of the measurement, a waiting time of 3 min was hold at the target temperatures.

3. Results and discussion

The XRD patterns of all the samples are shown in Fig. 1. For x = 0, the XRD pattern demonstrates a 4O modulated martensitic structure at RT. It suggests that the structural transformation in Ni48Co2Mn38Sn12 occurs above RT. While for x = 2.5, the presence of (111), (220), (400), and (422) peaks indicates a L21 austenitic structure with the lattice constant of a = 5.978 Å. The coexistence of austenitic and martenitic phases is observed in the samples with 0 < x < 2.5 at RT. The XRD data indicates that Tt decreases with x increasing.[24]

Fig. 1. (color online) Powder XRD patterns of Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys at RT.

The thermal-induced structural transformation was investigated by DSC measurements. As shown in Fig. 2, the large endothermic and exothermic peaks in the heating/cooling processes and the large thermal hysteresis indicate the occurrence of first-order structural transformation between martensite and austenite in these Heusler-type Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys. As x increases from 0 to 2.5, the Tt shifts to the lower temperatures by around 85 K and 110 K in the heating and cooling processes, respectively. It is known that the Tt of Ni–Mn-based shape memory alloys is sensitive to the electron concentration (e/a), and it decreases with the e/a decreasing.[2528] The e/a calculated from the EDS results for our samples are listed in Table 1. It can be found that with the increase of x, the e/a decreases from 8.115 to 8.025, resulting in the observed decrease of Tt.

Fig. 2. (color online) DSC curves for Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys in heating and cooling processes.
Table 1.

The measured compositions and the characteristic phase-transition temperatures of Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys.

.

Figure 3 shows the temperature dependence of magnetization (MT) of Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, 2.0, and 2.5) alloys under an applied field of 0.01 T during the heating and cooling processes. For x = 1.0, due to the fact of that the Tt in the heating process is close to the Curie-temperature of austenite (TC,A ≈ 340 K), the ferromagnetic austenite is not obviously observed in the heating process, but can be found in the cooling process. The existence of thermal hysteresis proves the first-order nature of the structural transformation between martensite and austenite. With x increasing from 1.5 to 2.5, the Tt shifts to the lower temperatures. As a result, the ferromagnetic austenite appears both in the heating and cooling processes. It is found that TC,A is nearly stable with the alteration of composition. The Curie temperature of martensite (TC,M) is not observed in the samples with x = 1.0, 1.5, and 2.0 in the temperature range of 250–400 K, but is found in the x = 2.5 sample. In the heating process, the sample with x = 2.5 experiences the magnetic ordering-disordering transition of martensite, and then the first-order MST from the low-magnetization martensite to ferromagnetic austenite. The TC,M of x = 2.5 is around 208 K. A phase diagram including austenitic start temperature As, austenitic finish temperature Af, martensitic start temperature Ms, martensitic finish temperature Mf, TC,M, and TC,A for Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys is shown in Fig. 4(a). According to Fig. 4(a), it can be found that with the increase of x, the Tt is decreased into the temperature window between TC,A and TC,M. Thus, the samples with x = 1.5 and 2.0 exhibit the room-temperature MST between the weak-magnetization martensite and ferromagnetic austenite. This transition is accompanied by a large ΔM, which is favor of MFIST. Due to the existence of ΔM between martensite and austenite, the MST can be induced by a magnetic field. The MB curves for Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, and 2.0) alloys with a temperature interval of 3 K in the heating process are shown in Figs. 4(b)4(d). With the increase of the magnetic field, an obvious slope change is observed when the temperature is close to As. This behavior suggests the existence of MFIST from the martensite to austenite.[2931]

Fig. 3. (color online) MT curves for Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, 2.0, and 2.5) alloys under an applied magnetic field of 0.01 T during the heating and cooling processes.
Fig. 4. (color online) (a) Phase diagram for Ni48−xCo2Mn38+xSn12 (x = 0, 1.0, 1.5, 2.0, and 2.5) alloys as a function of x. (b)-(d) MB curves for Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, and 2.0) alloys with a maximum magnetic field of 5 T.

According to the MB curves, the MCE of Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5 and 2.0) alloys is estimated by the Maxwell equation[32]

As shown in Fig. 5, the maximum values of magnetic entropy change (ΔSM) with the magnetic field variation of 0–5 T are 19.22 J⋅kg−1⋅K−1, 13.64 J⋅kg−1⋅K−1, and 7.71 J⋅kg−1⋅K−1 for x = 1.0, 1.5 and 2.0, respectively, and with the magnetic field variation of 0–2 T, the corresponding values decrease to 5.75 J⋅kg−1⋅K−1, 5.54 J⋅kg−1⋅K−1, and 3.57 J⋅kg−1⋅K−1, respectively.[3338] The maximum ΔSM decreases with the increase of x. It is found that the ΔSM is inversely proportional to the transformation width (W), which gradually increases with x increasing (shown in Table 2). As mentioned by Cugini et al.,[39] the transformation width is highly related to the structural discontinuity between the martensite and austenite phases, impurities, vacancies, and interstitial defects in Heusler-type alloys. Wu et al.[40] also reported that the magnetostructural transformation in the Mn–Ni–Ge alloy becomes broadened when the grain size decreases. Therefore, there are various factors that can influence the transformation width. In this work, it is found that accompanied by the increase of x, the transformation width increases. So, the observed broadened transformation width may be related to the increased chemical disorder. Besides ΔSM, the effective refrigeration capacity (RCeff) is also an important parameter to evaluate the magnetic cooling capacity of MCE materials. Here, RCeff is calculated by
where Tcold and Thot are the temperatures corresponding to the full width at half maximum (FWHM) of the ΔSM peak, and average HL is calculated from the area surrounded by the hysteresis loops (MB curves in Fig. 4), as shown in the inset of Fig. 5.[41,42] The calculated values for the transformation width, Tcold, Thot, ΔSM, average HL, and RCeff of Ni48−xCo2Mn38+xSn12 alloys are summarized in Table 2. Although the maximum ΔSM decreases with x increasing, the average HL is remarkably reduced. Therefore, an enhanced RCeff is obtained in the samples with x = 1.5 and 2. The calculated values of RCeff are 23.36 J⋅kg−1, 60.79 J⋅kg−1, and 67.60 J⋅kg−1 for x = 1.0, 1.5, and 2.0 with the magnetic field variation of 0–5 T, respectively, which are comparable to those of some other Ni–Mn-based Heusler alloys, such as Ni42Co8Mn30Fe2Ga18 (70 J⋅kg−1, ΔB = 5 T), Ni42.8Mn40.3Co5.7Sn11.2 (72.10 J⋅kg−1, ΔB = 3 T), Ni2Mn0.75Cu0.25Ga (72 J⋅kg−1, ΔB = 5 T), Ni50Mn37Sn13 (54 J⋅kg−1, ΔB = 5 T), Ni45Co5Mn38Sb9Ge3 (54 J⋅kg−1, ΔB = 5 T), and Ni52Mn26Ga22 (58.7 J⋅kg−1, ΔB = 5 T).[4348]

Fig. 5. (color online) Temperature dependence of magnetic entropy change for Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, and 2.0) alloys under the magnetic field variations of 0–2 and 0–5 T. The inset shows the x dependence of average HL for Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5 and 2.0) alloys under a magnetic field variation of 0–5 T.
Table 2.

The transformation width W(=AfAs), Tcold, Thot, magnetic entropy change ΔSM, hysteresis loss (HL), and effective refrigeration capacity (RCeff) of Ni48−xCo2Mn38+xSn12 (x = 1.0, 1.5, and 2.0) alloys.

.
4. Conclusion

In summary, we investigate the structural transformation and MCE in Ni48−xCo2Mn38+xSn12 alloys with x = 0, 1.0, 1.5, 2.0, and 2.5. The partial substitution of Ni by Mn stabilizes the austenite phase and leads to the decrease of Tt. For the samples with x = 1.5 and 2, the MST between the weak-magnetization martensite and ferromagnetic austenite is achieved at RT. This transition is accompanied by a large ΔM, which is favor of MFIST. Accompanied by the MFIST, a considerable MCE is obtained. Maximum values of ΔSM are 19.22 J⋅kg−1⋅K−1, 13.64 J⋅kg−1⋅K−1, and 7.71 J⋅kg−1⋅K− 1 for x = 1.0, 1.5, and 2.0 under the magnetic field variation of 0–5 T, respectively. Although the maximum ΔSM is reduced with x increasing, the RCeff is remarkably improved.

Reference
[1]GutfleischOWillardM ABruckEChenC HSankarS GLiuJ P 2011 Adv. Mater. 23 821
[2]GuillouFPorcariGYiboleHDijkVan NBruckE 2014 Adv. Mater. 26 2671
[3]LyubinaJ 2017 J. Phys. D: Appl. Phys. 50 053002
[4]ShahI AHassanNLiuJGongY YZhouX GFengX 2017 Chin. Phys. 26 017501
[5]LiuJGottschallTSkokovK PMooreJ DGutfleischO 2012 Nat. Mater. 11 620
[6]WadaHTanabeY 2001 Appl. Phys. Lett. 79 3302
[7]HuFShenBSunJWangGChengZ 2002 Appl. Phys. Lett. 80 826
[8]TagusOBruckEBuschowKBoerF D 2002 Nature 415 150
[9]HanZ DWangD HQianBFengJ FJiangX FDuY W 2010 J. Appl. Phys. 49 010211
[10]ZhangYLiuJZhengQZhangJXiaWDuJYaA 2014 Scr. Mater 75 26
[11]QuetzAKoshkidkoY STitovIRodionovIPandeySAryalAIbarra-GaytanP JPrudnikovVGranovskyADubenkoISamantaTCwikJLlamazaresJ L SStadlerSLahderantaEAliN 2016 J. Alloys Compd. 683 139
[12]XuanH CHanP DWangD HDuY W 2014 J. Alloys Compd. 582 369
[13]VarzanehA GKameliPKarimzadehFAslibeikiBVarvaroGSalamatiH 2014 J. Alloys Compd. 598 6
[14]LiZXuKZhangY LJingC 2015 J. Appl. Phys. 117 023902
[15]MaS CShihC WLiuJYuanJ HLeeS YLeeY IChangH WChangW C 2015 Acta Mater. 90 292
[16]XuanH CZhengY XMaS CCaoQ QWangD HDuY W 2010 J. Appl. Phys. 108 103920
[17]ChenF HHuangQ XJiangZ YXuanH CZhangM GXuX HZhaoJ W2016Smart Mater. Struct.256
[18]ZhengHWangWXueSZhaiQFrenzelJLuoZ 2013 Acta Mater 61 4648
[19]HuangLCongD YSuoH LWangY D 2014 Appl. Phys. Lett. 104 132407
[20]MaS CWangD HZhongZ CLuoJ MXuJ LDuY W 2013 Appl. Phys. Lett. 102 032407
[21]MoyaXNarayanS KMathurN D 2014 Nat. Mater. 13 439
[22]ChenFTongY XTianBLiLZhengY FLiuY 2013 J. Magn. Magn. Mater. 347 72
[23]CaronLOuZ QNguyenT TCam-ThanhD TTegusOBruckE 2009 J. Magn. Magn. Mater. 321 3559
[24]MaziarzWCzajaPSzczerbaM JDobrzynskaL LCzeppeTDutkiewiczJ 2014 J. Alloys Compd. 615 S173
[25]CongD YRothSSchultzL 2012 Acta Mater. 60 5335
[26]SrivastavaVChenXJamesR D 2010 Appl. Phys. Lett. 97 014101
[27]BhattiK PKhatibS ESrivastavaVJamesR DLeightonC 2012 Phys. Rev. 85 134450
[28]ChenF HGongC WGuoY PZhangM GChaiY S 2013 Phys. Status Solidi 210 2762
[29]EmreBBrunoN MEmreS YKaramanI 2014 Appl. Phys. Lett. 105 231910
[30]LuoHMengFJiangQLiuHLiuEWuGWangY 2010 Scr. Mater. 63 569
[31]YeMKimuraAMiuraYShiraiMCuiY TShimadaKNamatameHTaniguchiMUedaSKobayashiKKainumaRShishidoTFukushimaKKanomataT 2010 Phys. Rev. Lett. 104 176401
[32]ZhongZ CMaS CWangD HDuY W 2012 J. Mater. Sci. Technol. 28 193
[33]PecharskyV KGschneidnerK AJr 1997 Phys. Rev. Lett. 78 4494
[34]BenfordS MBrownG V 1981 J. Appl. Phys. 52 2110
[35]KrenkeTDumanEAcetMMoyaXManosaLPlanesA 2007 J. Appl. Phys. 102 033903
[36]SharmaV KChattopadhyayM KRoyS B 2010 J. Phys. D: Appl. Phys. 43 225001
[37]XuanH CChenF HHanP DWangD HDuY W 2014 Intermetallics 47 31
[38]NiemannRHeczkoOSchultzLFahlerS 2010 Appl. Phys. Lett. 97 222507
[39]Cugini1FPorcariGFabbriciSAlbertiniFSolziM 2016 Phil. Trans. R. Soc. 374 20150306
[40]WuRShenFHuFWangJBaoLZhangLLiuYZhaoYLiangFZuoWSunJShenB 2016 Sci. Rep. 6 20993
[41]BrockJKhanM 2017 J. Magn. Magn. Mater. 425 1
[42]SharmaV KChattopadhyayM KRoyS B 2007 J. Phys. D: Appl. Phys. 40 1869
[43]PathakA KDubenkoIKaracaH EStadlerSAliN 2010 Appl. Phys. Lett. 97 062505
[44]ChenF HGongC WGuoY PZhangM GChaiY S 2014 Chin. Phys. 23 067501
[45]StadlerSKhanMMitchellJAliNGomesA MDubenkoITakeuchiA YGuimaraesA P 2006 Appl. Phys. Lett. 88 192511
[46]PhanT LZhangPDanN HYenN HThanhP TThanhT DPhanM HYuS C 2012 Appl. Phys. Lett. 101 212403
[47]SahooRNayakA KSureshK GNigamA K2012J. Magn. Magn. Mater.3241267
[48]LiZ BZhangY DSanchez-ValdesC FSanchez-LlamazaresJ FEslingCZhaoXZuoL 2014 Appl. Phys. Lett. 104 044101