- Email: [email protected]
Magnetocaloric properties of cluster glass compound Pr2Ni0.95Si2.95
Intermetallics 111 (2019) 106490
Contents lists available at ScienceDirect
Intermetallics journal homepage: www.elsevier.com/locate/intermet
Magnetocaloric properties of cluster glass compound Pr2Ni0.95Si2.95 a,b
Santanu Pakhira a b
a,∗
a
, Chandan Mazumdar , R. Ranganathan
T
Condensed Matter Physics Division, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata, 700064, India Ames Laboratory-USDOE and Department of Physics and Astronomy, Iowa State University, Ames, IA, 50011, USA
ARTICLE INFO
ABSTRACT
Keywords: Intermetallics Rare-earth compounds X-ray diffraction Magnetization Magnetic entropy change Magnetocaloric effect
In this work, we report the magnetocaloric properties of a hexagonal intermetallic compound Pr2Ni0.95Si2.95 that exhibits spin cluster glass behaviour below 3.3 K. At lower temperature, the compound undergoes magnetic field induced metamagnetic transition beyond 5 kOe applied field value as inferred from the temperature dependence of magnetization at various fields as well as isothermal magnetization measurements. The temperature dependence of magnetic entropy change ( SM) has been calculated from the magnetization data and the maximum value is found to be 7.9 J/kg-K for a magnetic field change of 70 kOe. Interestingly, although the compound does not undergo any long range magnetic ordering, it exhibits considerable magnetocaloric effect in a temperature region much above the spin freezing due to the presence of short range magnetic correlation that persists up to much higher temperature. The suppression of short range interaction with the application of magnetic field has been argued to be responsible for such magnetic entropy change and adiabatic temperature change over a large temperature scale.
1. Introduction Magnetocaloric effect (MCE) based solid state refrigeration technique has been an exciting and active research field during the last decades in search for environment-friendly and cost effective replacement of gas refrigeration [1–6]. MCE is the ability of a material to change its temperature on the application of external magnetic field under adiabatic condition and generally quantified by isothermal entropy change (ΔSM) and adiabatic temperature change (ΔTad) of the material. Development in this research field relies on the finding of new materials that are potential candidates to be used as magnetic refrigerant. Some ferromagnetic and antiferromagnetic materials have already been reported to achieve ultralow temperatures with considerable relative cooling power (RCP) in research laboratory and are expected to make an impact in relevant industry also [7–10]. Among the different kinds of good magnetic refrigerant materials, rare-earth based magnetic systems are one of the promising set of candidates due to their large magnetocaloric properties arising from the localized 4f shell electrons [11–13]. There are quite a few number of such candidates with high moment rare-earth ions as constituent element. Although many compounds having high moment rare-earth ions are generally considered for their potentially good MCE behaviour, there exists only a very few compounds with low moment rare-earth ions. Among different rare-earth based systems, R2TSi3-type of
∗
compounds, with R = rare-earth ions, T = transition metals, are one of those recently found to exhibit large MCE [10,14–16]. These systems have generated great interest due to the interplay between crystal structure and ground state physical properties. Previous studies yield that most of the R2NiSi3-type of materials form in single phase only with defect crystal structure and also exhibit considerable MCE [17–21]. The study of MCE has also been used as an important tool to understand complex magnetic interaction in such systems [22]. Recently, it is reported that the Pr-based analogue forms in single phase with nominal composition Pr2Ni0.95Si2.95 (Fig. 1) and exhibits spin freezing behaviour below 3.3 K (Tf) in the absence of any true long range magnetic ordering [23]. In this work, we report detailed studies on magnetocaloric properties of this cluster-glass material. 2. Experimental details The polycrystalline Pr2Ni0.95Si2.95 was synthesized by melting appropriate amount of Pr (metal ingot, phase purity 99.9%), Ni (metal slug, phase purity 99.98%) and Si (metal lump, phase purity 99.9999%) in an arc furnace using a water-cooled copper hearth under inert (Ar) atmosphere. The sample was melted six times and after every melting the ingot was flipped to ensure homogeneity. The structural characterization was performed by Rietveld analysis of the powder x-ray diffraction (XRD) data taken on a TTRAX-III diffractometer (M/s
Corresponding author. E-mail addresses: [email protected], [email protected] (S. Pakhira), [email protected] (C. Mazumdar).
https://doi.org/10.1016/j.intermet.2019.106490 Received 23 October 2018; Received in revised form 21 April 2019; Accepted 26 April 2019 Available online 13 May 2019 0966-9795/ © 2019 Elsevier Ltd. All rights reserved.
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Fig. 1. Full Rietveld refinement of room temperature XRD data of Pr2Ni0.95Si2.95.
Rigaku, Japan) using Cu-Kα radiation. FULLPROF software package [24] was used to analyse the XRD data. The magnetic measurements were carried out in a superconducting quantum interference device (SQUID) (M/s Quantum Design Inc., USA) in the temperature range 2–300 K and magnetic fields up to 70 kOe. The heat capacity data were collected in a commercial Physical Properties Measurement System (PPMS) (M/s Quantum Design Inc., USA).
Fig. 2. (a) M(T) at different applied magnetic fields under ZFC and FC conditions. (b) Temperature derivative of ZFC magnetization as a function of temperature for different H.
strength beyond 5 kOe. To reveal the field evolution of the spin arrangements further, the isothermal magnetization measurements have been carried out at different temperatures and illustrated in Fig. 3(a). The M(H) behaviour at lowest temperature (2 K) is linear in low magnetic field region, while a tendency towards saturation is exhibited at higher field indicating a metamagnetic like transition. However, the magnetic moment value at 2 K for 70 kOe applied magnetic field strength found to be only 1.55 µ B which is much lower than the theoretical saturation moment value for Pr3+ ion (3.2 µ B ). No magnetic hysteresis is observed down to the lowest measured temperature. The exact strength of the magnetic field responsible for the metamagnetic transition in this system is determined from the dM/dH curve and found to be 5 kOe at 2 K (Inset: Fig. 3(a)). As seen from Fig. 3(a), the linear M(H) behaviour is only observed for temperatures higher than 25 K. As reported earlier, this is due to the presence of short range magnetic correlations persisting up to such higher temperature [23]. In a magnetic system, Arrott plots (M2 vs. H/M) are generally used to determine the phase transitions temperature and the nature of long range magnetic phase transition. According to the Banerjee criterion [28], a positive slope is obtained for a second order phase transition, while a first order phase transition is identified by a negative slope of M2(H/M) curve. In cluster glass system Pr2Ni0.95Si2.95, where the quench disorder restricts the magnetic coherence length, it would not be wise to determine the order of transition from such analysis. However, from the M2(H/M) behaviour, it is established that no signature of spontaneous magnetization could be observed in the system (Fig. 3(b)). Instead, the intercept on the M2 axis obtained by extrapolation from the high field region confirms the presence of a metamagnetic transition at low temperatures. As the magnetically frustrated spin freezing states are associated with large degeneracy, such magnetic field induced polarization is expected to release a considerable amount of magnetic entropy. This encourages us to study the magnetocaloric properties of Pr2Ni0.95Si2.95 by calculating the magnetic entropy changes of the compound for different applied magnetic fields. The isothermal magnetic entropy changes (ΔSM) for Pr2Ni0.95Si2.95 were calculated from a set of magnetic isotherms shown in Fig. 3(c) within temperature range 2 K–32 K and magnetic fields up to 70 kOe
3. Results and discussions Our attempt to synthesize stoichiometric Pr2NiSi3 resulted the formation of an additional secondary phase PrNiSi2, as reported earlier [23]. The presence of such secondary phase has earlier been reported in many members of R2NiSi3 series of materials [17–20,25,26], except for R = Gd & Er [10]. The single phase compound can however be synthesized by deliberately introducing elemental vacancies of appropriate proportion in Pr2Ni0.95Si2.95, although a trace of unreacted elemental Pr could still be found in the XRD pattern at 2θ ∼ 29.9∘ (Fig. 1). The room temperature XRD pattern of Pr2Ni0.95Si2.95 along with full Rietveld refinement is shown in Fig. 1. The lattice parameters of the essentially single phase compound, as estimated from the XRD data taken at room temperature, are found to be are a = 4.032(1) Å and c = 4.247(1) Å (space group P6/mmm). The energy dispersive X-ray spectroscopy (EDX) analysis indicates the average composition to be Pr2Ni0.89Si2.99 [23]. The temperature dependence of magnetization at different applied magnetic fields for Pr2Ni0.95Si2.95 are shown in Fig. 2 (a). As seen from the figure, the magnetic transition peak is not as sharp as generally expected for long range antiferromagnetic order, but spread over a finite temperature region. Such cusp in the dc magnetic susceptibility data along with the thermomagnetic irreversibility in zero-field-cooled (ZFC) and field-cooled (FC) magnetization behaviour hint towards a metastable state formation in the system. The ac susceptibility data confirms the appearance of cluster glass state below 3.3 K in Pr2Ni0.95Si2.95 [23]. Though a spin freezing cusp appears in the M(T) behaviour for low applied magnetic field value, the peak smears out for magnetic field beyond 5 kOe. This behaviour is better reflected in the temperature derivative of magnetization behaviour for different magnetic fields as shown in Fig. 2(b). The dM/dt behaviour for H < 5 kOe shows a crossover from positive to negative value. However for H ≥ 5 kOe, the dM/dT behaviour only exhibits a dip in the temperature axis similar to that found in a ferromagnetic system. The irreversibility between ZFC and FC magnetization which is generally considered to be a key signature for spin freezing ground state [27], is also vanished for H ≥ 1 kOe. These results clearly demonstrate that the ground state spin configuration tends to polarize in the magnetic field direction resulting a net ferromagnetic component in the system for magnetic field 2
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Fig. 3. (a) M(H) at different temperatures. Inset shows dM/dH for T = 2 K. (b) Arrott plots (M2 vs. H/M) at different temperatures of Pr2Ni0.95Si2.95. (c) M(H) at different temperatures for magnetic field change 0 → 70 kOe.
Fig. 4. (a) Isothermal magnetic entropy change ( SM) as a function of temperature for different magnetic fields. (b) SM(ΔH) along with H2 dependence for some selected temperatures. (c) RCP as a function of ΔH.
using Maxwell's relation,
SM =
H2 H1
M dH T
symmetric around the maximum position. Such type of asymmetric spread SM(T) curve is previously reported for materials having spin flucof tuation up to much higher temperature than the phase transition temSM follows a Cm H 2/2T 2 perature [10,31–33]. It is well-established that, like behaviour in the paramagnetic region, with Cm the molar Curie conSM(H) behaviour stant. Fig. 4(b) shows the experimentally estimated for Pr2Ni0.95Si2.95 at some selected temperatures along with the theoretically calculated H2 dependence. As seen from the figure, the behaviour of SM(H) significantly deviates from the H2 dependence for 9.5 K, 15 K and 22 K, which are well above the freezing temperature in this compound. The fit to H2 dependence improves significantly for T ≥ 30 K. This result is in accordance with the magnetic susceptibility data that reveals that true paramagnetic behaviour in this compound is only recovered above 30 K [23]. Thus, the short-range correlation that dominates up to temperatures well above Tf results in a large magnetic entropy change under application of an external magnetic field at such intermediate temperatures. Due to this asymmetric broadening of SM(T), the relative cooling power (RCP), defined as RCP = SmaxM × δTFWHM, where δTFWHM is the full width at half maximum of SM(T) curve, is also found to be considerable. RCP is another significant parameter that quantifies in a magnetic material the amount of heat transfer between the hot and cold sinks in an ideal
(1)
Fig. 4(a) depicts the temperature dependence of SM of the compound for different magnetic field changes. The maximum value of magnetic entropy change ( SmaxM) is found to be 7.9 J/kg-K for a field change of 70 kOe. This moderate value is comparable to that reported for both low and high moment rare-earth based good magnetocaloric materials in the cryogenic temperature region, viz., Gd2Co3Al9( SmaxM = 6.63 J/ max kg-K for ΔH = 70 kOe) [12], PrNi ( S M = 6.1 J/kg-K for ΔH = 50 kOe) [29], Pr2CuSi3( SmaxM = 7.6 J/kg-K for ΔH = 50 kOe) [16], Er3Pd2( SmaxM = 8.9 J/kg-K for ΔH = 50 kOe) [30], etc. It may be pointed out here that the maximum value of magnetic entropy change for this compound is achieved at a temperature region higher than the spin freezing temperature where short range magnetic correlation dominates. One may also note that SmaxM(H) does not show any saturation tendency, but keeps on increasing monotonically up to the highest applied magnetic field strength of 70 kOe. It may be due to fact that a much higher magnetic field is required to melt the spin freezed state and to polarize fully in the magnetic field direction. The SM(T) behaviour is not found to be 3
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Acknowledgement This work has been carried out and supported through CMPID-DAE project at SINP. SP acknowledges the receipt of fellowship from Ames Laboratory-USDOE and Department of Physics and Astronomy, Iowa State University, USA. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.intermet.2019.106490. References Fig. 5. Temperature dependence of adiabatic temperature change (ΔTad) for different magnetic fields. Inset shows low temperature region of zero field heat capacity data for Pr2Ni0.95Si2.95.
[1] V.K. Pecharsky, K.A. Gschneidner Jr., Phys. Rev. Lett. 78 (1997) 4494. [2] M.H. Phan, S.C. Yu, J. Magn. Magn. Mater. 308 (2007) 325. [3] A.M. Tishin, Y.I. Spichkin, The Magnetocaloric Effect and its Applications, IOP, New York, 2003. [4] B.F. Yu, Q. Gao, B. Zhang, X.Z. Meng, Z. Chen, Int. J. Refrig. 26 (2003) 622. [5] L. Zhang, S.A. Sherif, A.J. Degregoria, C.B. Zimm, T.N. Veziroglu, Cryogenics 40 (2000) 269. [6] K.A. Gschneidner Jr., V.K. Pecharsky, A.O. Tsokol, Rep. Prog. Phys. 68 (2005) 1479. [7] O. Tegus, E. Brück, K.H.J. Buschow, F.R. de Boer, Nature(London) 415 (2002) 150. [8] V. Provenzano, A.J. Shapiro, R.D. Shull, Nature(London) 429 (2004) 853. [9] T. Samanta, I. Das, S. Banerjee, Appl. Phys. Lett. 91 (2007) 152506. [10] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, M. Avdeev, Phys. Rev. B 94 (2016) 104414. [11] N.K. Singh, K.G. Suresh, A.K. Nigam, S.K. Malik, A.A. Coelho, S. Gama, J. Magn. Magn. Mater. 317 (2007) 68. [12] T.P. Rashid, K. Arun, I. Curlik, S. Ilkovic, M. Reiffers, A. Dzubinska, R. Nagalakshmi, J. Magn. Magn. Mater. 466 (2018) 283. [13] Mohit K. Sharma, K. Mukherjee, J. Magn. Magn. Mater. 444 (2017) 178. [14] Z.J. Mo, J. Shen, L.Q. Yan, X.Q. Gao, C.C. Tang, J.F. Wu, J.R. Sun, B.G. Shen, J. Alloy. Comp. 618 (2015) 512. [15] Z.J. Mo, J. Shen, X.Q. Gao, Y. Liu, C.C. Tang, J.F. Wu, F.X. Hu, J.R. Sun, B.G. Shen, J. Alloy. Comp. 626 (2015) 145. [16] F. Wang, Feng-ying Yuan, Jin-zhi Wang, Tang-fu Feng, Guo-qi Hu, J. Alloy. Comp. 592 (2014) 63. [17] S. Pakhira, C. Mazumdar, R. Ranganathan, M. Avdeev, Sci. Rep. 7 (2017) 7367. [18] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Phys. Condens. Matter 29 (2017) 505801. [19] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, Phys. Chem. Chem. Phys. 20 (2018) 7082. [20] S. Pakhira, C. Mazumdar, D. Choudhury, R. Ranganathan, S. Giri, Phys. Chem. Chem. Phys. 20 (2018) 13580. [21] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Magn. Magn. Mater. 484 (2019) 456, https://doi.org/10.1016/j.jmmm.2019.04.021. [22] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Phys. Condens. Matter 30 (2018) 215601. [23] S. Pakhira, C. Mazumdar, A. Basu, R. Ranganathan, R.N. Bhowmik, B. Satpati, Sci. Rep. 8 (2018) 14870. [24] Rodríguez-Carvajal, J. Phys. B 192 (1993) 55. [25] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, J. Alloy. Comp. 742 (2018) 391. [26] S. Pakhira, C. Mazumdar, M. Avdeev, R.N. Bhowmik, R. Ranganathan, J. Alloy. Comp. 785 (2019) 72. [27] J.A. Mydosh, Spin Glasses: an Experimental Introduction, Taylor & Francis, London, Washington, 1993. [28] S.K. Banerjee, Phys. Lett. 12 (1964) 16. [29] A.O. Pecharsky, Yu Mozharivskyj, K.W. Dennis, K.A. Gschneidner Jr., R.W. McCallum, G.J. Miller, V.K. Pecharsky, Phys. Rev. B 68 (2003) 134452. [30] B. Maji, M.K. Ray, M. Modak, S. Mondal, K.G. Suresh, S. Banerjee, J. Magn. Magn. Mater. 456 (2018) 236. [31] S. Gupta, R. Rawat, K.G. Suresh, Appl. Phys. Lett. 105 (2014) 012403. [32] S.K. Tripathy, K.G. Suresh, A.K. Nigam, J. Magn. Magn. Mater. 306 (2006) 24. [33] R. Rawat, I. Das, J. Phys. Condens. Matter 13 (2001) L379. [34] L. Li, K. Nishimura, M. Kadonaga, Z.H. Qian, D.X. Huo, J. Appl. Phys. 110 (2011) 043912. [35] Y. Zhang, Y. Yang, X. Xu, S. Geng, L. Hou, X. Li, Z. Ren, G. Wilde, Sci. Rep. 6 (2016) 34192. [36] L. Li, O. Niehaus, B. Gerke, R. Pöttgen, IEEE Trans. Magn. 50 (2014) 2503604. [37] L. Li, O. Niehaus, M. Kersting, R. Pöttgen, Intermetallics 62 (2015) 17. [38] B. Maji, K.G. Suresh, A.K. Nigam, Appl. Phys. Lett. 102 (2013) 062406.
refrigeration cycle. The maximum value of RCP is estimated in this case to be 146 J/kg for 70 kOe magnetic field change that is in the same range to that reported for Gd2Co3Al9 (201 J/kg-K for ΔH = 70 kOe) [12], ErNi2B2C (155 J/kg-K for ΔH = 50 kOe) [34], Tm2Cu2Cd (165 J/kg-K for ΔH = 70 kOe) [35], etc for cryogenic temperatures. The adiabatic temperature change (ΔTad), another important parameter to describe the performance of a good magnetocaloric material, SM(T) and zero-field heat capacity data has been estimated using (Inset: Fig. 5) using the relation,
Tad = [T (S, H )
T (S, 0)]S
(2)
The temperature dependence of ΔTad for different magnetic fields is shown in Fig. 5. The maximum value of ΔTad has been found to be 3.85 K for 70 kOe magnetic field change and in the same range to that reported for Pr5Ni1.9Si3 (3.4 K for ΔH = 75 kOe) [29], EuAuZn (3.8 K for ΔH = 50 kOe) [36], Ho4PtMg (5.3 K for ΔH = 70 kOe) [37], NdRu2Ge2(2.7 K for ΔH = 50 kOe) [38], etc in the low temperature region. Similar to SM(T), the behaviour of ΔTad(T) is also asymmetric over the maximum value and the maximum change is also observed at much higher temperature than the spin freezing temperature for the compound. Thus, we have shown that the presence of short range correlation above the spin freezing temperature results in a considerable magnetocaloric effect in Pr2Ni0.95Si2.95. 3.1. Summary In summary, we have studied the magnetocaloric properties of the magnetically frustrated cluster glass material Pr2Ni0.95Si2.95. The ground state of the material is found to be quite fragile and undergoes a metamagnetic transition at magnetic fields above 5 kOe at 2K. This results in moderate magnetic entropy change ( SM) of 7.9 J/kg-K, adiabatic temperature change (ΔTad) of 3.85 K along with RCP value of 146 J/kg for a field change of 70 kOe. The short range magnetic correlation spread over a much higher temperature region than the spin freezing temperature is responsible for the achievement of substantial SM and ΔTad values with asymmetric temperature dependencies. This is one of those few rare-earth based compounds that exhibit considerable magnetocaloric effect despite having low moment rare-earth ion. We believe that these results establish Pr2Ni0.95Si2.95 to be an interesting magnetic material not only for its possible technological application but for fundamental understanding as well.
4
Contents lists available at ScienceDirect
Intermetallics journal homepage: www.elsevier.com/locate/intermet
Magnetocaloric properties of cluster glass compound Pr2Ni0.95Si2.95 a,b
Santanu Pakhira a b
a,∗
a
, Chandan Mazumdar , R. Ranganathan
T
Condensed Matter Physics Division, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata, 700064, India Ames Laboratory-USDOE and Department of Physics and Astronomy, Iowa State University, Ames, IA, 50011, USA
ARTICLE INFO
ABSTRACT
Keywords: Intermetallics Rare-earth compounds X-ray diffraction Magnetization Magnetic entropy change Magnetocaloric effect
In this work, we report the magnetocaloric properties of a hexagonal intermetallic compound Pr2Ni0.95Si2.95 that exhibits spin cluster glass behaviour below 3.3 K. At lower temperature, the compound undergoes magnetic field induced metamagnetic transition beyond 5 kOe applied field value as inferred from the temperature dependence of magnetization at various fields as well as isothermal magnetization measurements. The temperature dependence of magnetic entropy change ( SM) has been calculated from the magnetization data and the maximum value is found to be 7.9 J/kg-K for a magnetic field change of 70 kOe. Interestingly, although the compound does not undergo any long range magnetic ordering, it exhibits considerable magnetocaloric effect in a temperature region much above the spin freezing due to the presence of short range magnetic correlation that persists up to much higher temperature. The suppression of short range interaction with the application of magnetic field has been argued to be responsible for such magnetic entropy change and adiabatic temperature change over a large temperature scale.
1. Introduction Magnetocaloric effect (MCE) based solid state refrigeration technique has been an exciting and active research field during the last decades in search for environment-friendly and cost effective replacement of gas refrigeration [1–6]. MCE is the ability of a material to change its temperature on the application of external magnetic field under adiabatic condition and generally quantified by isothermal entropy change (ΔSM) and adiabatic temperature change (ΔTad) of the material. Development in this research field relies on the finding of new materials that are potential candidates to be used as magnetic refrigerant. Some ferromagnetic and antiferromagnetic materials have already been reported to achieve ultralow temperatures with considerable relative cooling power (RCP) in research laboratory and are expected to make an impact in relevant industry also [7–10]. Among the different kinds of good magnetic refrigerant materials, rare-earth based magnetic systems are one of the promising set of candidates due to their large magnetocaloric properties arising from the localized 4f shell electrons [11–13]. There are quite a few number of such candidates with high moment rare-earth ions as constituent element. Although many compounds having high moment rare-earth ions are generally considered for their potentially good MCE behaviour, there exists only a very few compounds with low moment rare-earth ions. Among different rare-earth based systems, R2TSi3-type of
∗
compounds, with R = rare-earth ions, T = transition metals, are one of those recently found to exhibit large MCE [10,14–16]. These systems have generated great interest due to the interplay between crystal structure and ground state physical properties. Previous studies yield that most of the R2NiSi3-type of materials form in single phase only with defect crystal structure and also exhibit considerable MCE [17–21]. The study of MCE has also been used as an important tool to understand complex magnetic interaction in such systems [22]. Recently, it is reported that the Pr-based analogue forms in single phase with nominal composition Pr2Ni0.95Si2.95 (Fig. 1) and exhibits spin freezing behaviour below 3.3 K (Tf) in the absence of any true long range magnetic ordering [23]. In this work, we report detailed studies on magnetocaloric properties of this cluster-glass material. 2. Experimental details The polycrystalline Pr2Ni0.95Si2.95 was synthesized by melting appropriate amount of Pr (metal ingot, phase purity 99.9%), Ni (metal slug, phase purity 99.98%) and Si (metal lump, phase purity 99.9999%) in an arc furnace using a water-cooled copper hearth under inert (Ar) atmosphere. The sample was melted six times and after every melting the ingot was flipped to ensure homogeneity. The structural characterization was performed by Rietveld analysis of the powder x-ray diffraction (XRD) data taken on a TTRAX-III diffractometer (M/s
Corresponding author. E-mail addresses: [email protected], [email protected] (S. Pakhira), [email protected] (C. Mazumdar).
https://doi.org/10.1016/j.intermet.2019.106490 Received 23 October 2018; Received in revised form 21 April 2019; Accepted 26 April 2019 Available online 13 May 2019 0966-9795/ © 2019 Elsevier Ltd. All rights reserved.
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Fig. 1. Full Rietveld refinement of room temperature XRD data of Pr2Ni0.95Si2.95.
Rigaku, Japan) using Cu-Kα radiation. FULLPROF software package [24] was used to analyse the XRD data. The magnetic measurements were carried out in a superconducting quantum interference device (SQUID) (M/s Quantum Design Inc., USA) in the temperature range 2–300 K and magnetic fields up to 70 kOe. The heat capacity data were collected in a commercial Physical Properties Measurement System (PPMS) (M/s Quantum Design Inc., USA).
Fig. 2. (a) M(T) at different applied magnetic fields under ZFC and FC conditions. (b) Temperature derivative of ZFC magnetization as a function of temperature for different H.
strength beyond 5 kOe. To reveal the field evolution of the spin arrangements further, the isothermal magnetization measurements have been carried out at different temperatures and illustrated in Fig. 3(a). The M(H) behaviour at lowest temperature (2 K) is linear in low magnetic field region, while a tendency towards saturation is exhibited at higher field indicating a metamagnetic like transition. However, the magnetic moment value at 2 K for 70 kOe applied magnetic field strength found to be only 1.55 µ B which is much lower than the theoretical saturation moment value for Pr3+ ion (3.2 µ B ). No magnetic hysteresis is observed down to the lowest measured temperature. The exact strength of the magnetic field responsible for the metamagnetic transition in this system is determined from the dM/dH curve and found to be 5 kOe at 2 K (Inset: Fig. 3(a)). As seen from Fig. 3(a), the linear M(H) behaviour is only observed for temperatures higher than 25 K. As reported earlier, this is due to the presence of short range magnetic correlations persisting up to such higher temperature [23]. In a magnetic system, Arrott plots (M2 vs. H/M) are generally used to determine the phase transitions temperature and the nature of long range magnetic phase transition. According to the Banerjee criterion [28], a positive slope is obtained for a second order phase transition, while a first order phase transition is identified by a negative slope of M2(H/M) curve. In cluster glass system Pr2Ni0.95Si2.95, where the quench disorder restricts the magnetic coherence length, it would not be wise to determine the order of transition from such analysis. However, from the M2(H/M) behaviour, it is established that no signature of spontaneous magnetization could be observed in the system (Fig. 3(b)). Instead, the intercept on the M2 axis obtained by extrapolation from the high field region confirms the presence of a metamagnetic transition at low temperatures. As the magnetically frustrated spin freezing states are associated with large degeneracy, such magnetic field induced polarization is expected to release a considerable amount of magnetic entropy. This encourages us to study the magnetocaloric properties of Pr2Ni0.95Si2.95 by calculating the magnetic entropy changes of the compound for different applied magnetic fields. The isothermal magnetic entropy changes (ΔSM) for Pr2Ni0.95Si2.95 were calculated from a set of magnetic isotherms shown in Fig. 3(c) within temperature range 2 K–32 K and magnetic fields up to 70 kOe
3. Results and discussions Our attempt to synthesize stoichiometric Pr2NiSi3 resulted the formation of an additional secondary phase PrNiSi2, as reported earlier [23]. The presence of such secondary phase has earlier been reported in many members of R2NiSi3 series of materials [17–20,25,26], except for R = Gd & Er [10]. The single phase compound can however be synthesized by deliberately introducing elemental vacancies of appropriate proportion in Pr2Ni0.95Si2.95, although a trace of unreacted elemental Pr could still be found in the XRD pattern at 2θ ∼ 29.9∘ (Fig. 1). The room temperature XRD pattern of Pr2Ni0.95Si2.95 along with full Rietveld refinement is shown in Fig. 1. The lattice parameters of the essentially single phase compound, as estimated from the XRD data taken at room temperature, are found to be are a = 4.032(1) Å and c = 4.247(1) Å (space group P6/mmm). The energy dispersive X-ray spectroscopy (EDX) analysis indicates the average composition to be Pr2Ni0.89Si2.99 [23]. The temperature dependence of magnetization at different applied magnetic fields for Pr2Ni0.95Si2.95 are shown in Fig. 2 (a). As seen from the figure, the magnetic transition peak is not as sharp as generally expected for long range antiferromagnetic order, but spread over a finite temperature region. Such cusp in the dc magnetic susceptibility data along with the thermomagnetic irreversibility in zero-field-cooled (ZFC) and field-cooled (FC) magnetization behaviour hint towards a metastable state formation in the system. The ac susceptibility data confirms the appearance of cluster glass state below 3.3 K in Pr2Ni0.95Si2.95 [23]. Though a spin freezing cusp appears in the M(T) behaviour for low applied magnetic field value, the peak smears out for magnetic field beyond 5 kOe. This behaviour is better reflected in the temperature derivative of magnetization behaviour for different magnetic fields as shown in Fig. 2(b). The dM/dt behaviour for H < 5 kOe shows a crossover from positive to negative value. However for H ≥ 5 kOe, the dM/dT behaviour only exhibits a dip in the temperature axis similar to that found in a ferromagnetic system. The irreversibility between ZFC and FC magnetization which is generally considered to be a key signature for spin freezing ground state [27], is also vanished for H ≥ 1 kOe. These results clearly demonstrate that the ground state spin configuration tends to polarize in the magnetic field direction resulting a net ferromagnetic component in the system for magnetic field 2
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Fig. 3. (a) M(H) at different temperatures. Inset shows dM/dH for T = 2 K. (b) Arrott plots (M2 vs. H/M) at different temperatures of Pr2Ni0.95Si2.95. (c) M(H) at different temperatures for magnetic field change 0 → 70 kOe.
Fig. 4. (a) Isothermal magnetic entropy change ( SM) as a function of temperature for different magnetic fields. (b) SM(ΔH) along with H2 dependence for some selected temperatures. (c) RCP as a function of ΔH.
using Maxwell's relation,
SM =
H2 H1
M dH T
symmetric around the maximum position. Such type of asymmetric spread SM(T) curve is previously reported for materials having spin flucof tuation up to much higher temperature than the phase transition temSM follows a Cm H 2/2T 2 perature [10,31–33]. It is well-established that, like behaviour in the paramagnetic region, with Cm the molar Curie conSM(H) behaviour stant. Fig. 4(b) shows the experimentally estimated for Pr2Ni0.95Si2.95 at some selected temperatures along with the theoretically calculated H2 dependence. As seen from the figure, the behaviour of SM(H) significantly deviates from the H2 dependence for 9.5 K, 15 K and 22 K, which are well above the freezing temperature in this compound. The fit to H2 dependence improves significantly for T ≥ 30 K. This result is in accordance with the magnetic susceptibility data that reveals that true paramagnetic behaviour in this compound is only recovered above 30 K [23]. Thus, the short-range correlation that dominates up to temperatures well above Tf results in a large magnetic entropy change under application of an external magnetic field at such intermediate temperatures. Due to this asymmetric broadening of SM(T), the relative cooling power (RCP), defined as RCP = SmaxM × δTFWHM, where δTFWHM is the full width at half maximum of SM(T) curve, is also found to be considerable. RCP is another significant parameter that quantifies in a magnetic material the amount of heat transfer between the hot and cold sinks in an ideal
(1)
Fig. 4(a) depicts the temperature dependence of SM of the compound for different magnetic field changes. The maximum value of magnetic entropy change ( SmaxM) is found to be 7.9 J/kg-K for a field change of 70 kOe. This moderate value is comparable to that reported for both low and high moment rare-earth based good magnetocaloric materials in the cryogenic temperature region, viz., Gd2Co3Al9( SmaxM = 6.63 J/ max kg-K for ΔH = 70 kOe) [12], PrNi ( S M = 6.1 J/kg-K for ΔH = 50 kOe) [29], Pr2CuSi3( SmaxM = 7.6 J/kg-K for ΔH = 50 kOe) [16], Er3Pd2( SmaxM = 8.9 J/kg-K for ΔH = 50 kOe) [30], etc. It may be pointed out here that the maximum value of magnetic entropy change for this compound is achieved at a temperature region higher than the spin freezing temperature where short range magnetic correlation dominates. One may also note that SmaxM(H) does not show any saturation tendency, but keeps on increasing monotonically up to the highest applied magnetic field strength of 70 kOe. It may be due to fact that a much higher magnetic field is required to melt the spin freezed state and to polarize fully in the magnetic field direction. The SM(T) behaviour is not found to be 3
Intermetallics 111 (2019) 106490
S. Pakhira, et al.
Acknowledgement This work has been carried out and supported through CMPID-DAE project at SINP. SP acknowledges the receipt of fellowship from Ames Laboratory-USDOE and Department of Physics and Astronomy, Iowa State University, USA. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.intermet.2019.106490. References Fig. 5. Temperature dependence of adiabatic temperature change (ΔTad) for different magnetic fields. Inset shows low temperature region of zero field heat capacity data for Pr2Ni0.95Si2.95.
[1] V.K. Pecharsky, K.A. Gschneidner Jr., Phys. Rev. Lett. 78 (1997) 4494. [2] M.H. Phan, S.C. Yu, J. Magn. Magn. Mater. 308 (2007) 325. [3] A.M. Tishin, Y.I. Spichkin, The Magnetocaloric Effect and its Applications, IOP, New York, 2003. [4] B.F. Yu, Q. Gao, B. Zhang, X.Z. Meng, Z. Chen, Int. J. Refrig. 26 (2003) 622. [5] L. Zhang, S.A. Sherif, A.J. Degregoria, C.B. Zimm, T.N. Veziroglu, Cryogenics 40 (2000) 269. [6] K.A. Gschneidner Jr., V.K. Pecharsky, A.O. Tsokol, Rep. Prog. Phys. 68 (2005) 1479. [7] O. Tegus, E. Brück, K.H.J. Buschow, F.R. de Boer, Nature(London) 415 (2002) 150. [8] V. Provenzano, A.J. Shapiro, R.D. Shull, Nature(London) 429 (2004) 853. [9] T. Samanta, I. Das, S. Banerjee, Appl. Phys. Lett. 91 (2007) 152506. [10] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, M. Avdeev, Phys. Rev. B 94 (2016) 104414. [11] N.K. Singh, K.G. Suresh, A.K. Nigam, S.K. Malik, A.A. Coelho, S. Gama, J. Magn. Magn. Mater. 317 (2007) 68. [12] T.P. Rashid, K. Arun, I. Curlik, S. Ilkovic, M. Reiffers, A. Dzubinska, R. Nagalakshmi, J. Magn. Magn. Mater. 466 (2018) 283. [13] Mohit K. Sharma, K. Mukherjee, J. Magn. Magn. Mater. 444 (2017) 178. [14] Z.J. Mo, J. Shen, L.Q. Yan, X.Q. Gao, C.C. Tang, J.F. Wu, J.R. Sun, B.G. Shen, J. Alloy. Comp. 618 (2015) 512. [15] Z.J. Mo, J. Shen, X.Q. Gao, Y. Liu, C.C. Tang, J.F. Wu, F.X. Hu, J.R. Sun, B.G. Shen, J. Alloy. Comp. 626 (2015) 145. [16] F. Wang, Feng-ying Yuan, Jin-zhi Wang, Tang-fu Feng, Guo-qi Hu, J. Alloy. Comp. 592 (2014) 63. [17] S. Pakhira, C. Mazumdar, R. Ranganathan, M. Avdeev, Sci. Rep. 7 (2017) 7367. [18] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Phys. Condens. Matter 29 (2017) 505801. [19] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, Phys. Chem. Chem. Phys. 20 (2018) 7082. [20] S. Pakhira, C. Mazumdar, D. Choudhury, R. Ranganathan, S. Giri, Phys. Chem. Chem. Phys. 20 (2018) 13580. [21] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Magn. Magn. Mater. 484 (2019) 456, https://doi.org/10.1016/j.jmmm.2019.04.021. [22] S. Pakhira, C. Mazumdar, R. Ranganathan, J. Phys. Condens. Matter 30 (2018) 215601. [23] S. Pakhira, C. Mazumdar, A. Basu, R. Ranganathan, R.N. Bhowmik, B. Satpati, Sci. Rep. 8 (2018) 14870. [24] Rodríguez-Carvajal, J. Phys. B 192 (1993) 55. [25] S. Pakhira, C. Mazumdar, R. Ranganathan, S. Giri, J. Alloy. Comp. 742 (2018) 391. [26] S. Pakhira, C. Mazumdar, M. Avdeev, R.N. Bhowmik, R. Ranganathan, J. Alloy. Comp. 785 (2019) 72. [27] J.A. Mydosh, Spin Glasses: an Experimental Introduction, Taylor & Francis, London, Washington, 1993. [28] S.K. Banerjee, Phys. Lett. 12 (1964) 16. [29] A.O. Pecharsky, Yu Mozharivskyj, K.W. Dennis, K.A. Gschneidner Jr., R.W. McCallum, G.J. Miller, V.K. Pecharsky, Phys. Rev. B 68 (2003) 134452. [30] B. Maji, M.K. Ray, M. Modak, S. Mondal, K.G. Suresh, S. Banerjee, J. Magn. Magn. Mater. 456 (2018) 236. [31] S. Gupta, R. Rawat, K.G. Suresh, Appl. Phys. Lett. 105 (2014) 012403. [32] S.K. Tripathy, K.G. Suresh, A.K. Nigam, J. Magn. Magn. Mater. 306 (2006) 24. [33] R. Rawat, I. Das, J. Phys. Condens. Matter 13 (2001) L379. [34] L. Li, K. Nishimura, M. Kadonaga, Z.H. Qian, D.X. Huo, J. Appl. Phys. 110 (2011) 043912. [35] Y. Zhang, Y. Yang, X. Xu, S. Geng, L. Hou, X. Li, Z. Ren, G. Wilde, Sci. Rep. 6 (2016) 34192. [36] L. Li, O. Niehaus, B. Gerke, R. Pöttgen, IEEE Trans. Magn. 50 (2014) 2503604. [37] L. Li, O. Niehaus, M. Kersting, R. Pöttgen, Intermetallics 62 (2015) 17. [38] B. Maji, K.G. Suresh, A.K. Nigam, Appl. Phys. Lett. 102 (2013) 062406.
refrigeration cycle. The maximum value of RCP is estimated in this case to be 146 J/kg for 70 kOe magnetic field change that is in the same range to that reported for Gd2Co3Al9 (201 J/kg-K for ΔH = 70 kOe) [12], ErNi2B2C (155 J/kg-K for ΔH = 50 kOe) [34], Tm2Cu2Cd (165 J/kg-K for ΔH = 70 kOe) [35], etc for cryogenic temperatures. The adiabatic temperature change (ΔTad), another important parameter to describe the performance of a good magnetocaloric material, SM(T) and zero-field heat capacity data has been estimated using (Inset: Fig. 5) using the relation,
Tad = [T (S, H )
T (S, 0)]S
(2)
The temperature dependence of ΔTad for different magnetic fields is shown in Fig. 5. The maximum value of ΔTad has been found to be 3.85 K for 70 kOe magnetic field change and in the same range to that reported for Pr5Ni1.9Si3 (3.4 K for ΔH = 75 kOe) [29], EuAuZn (3.8 K for ΔH = 50 kOe) [36], Ho4PtMg (5.3 K for ΔH = 70 kOe) [37], NdRu2Ge2(2.7 K for ΔH = 50 kOe) [38], etc in the low temperature region. Similar to SM(T), the behaviour of ΔTad(T) is also asymmetric over the maximum value and the maximum change is also observed at much higher temperature than the spin freezing temperature for the compound. Thus, we have shown that the presence of short range correlation above the spin freezing temperature results in a considerable magnetocaloric effect in Pr2Ni0.95Si2.95. 3.1. Summary In summary, we have studied the magnetocaloric properties of the magnetically frustrated cluster glass material Pr2Ni0.95Si2.95. The ground state of the material is found to be quite fragile and undergoes a metamagnetic transition at magnetic fields above 5 kOe at 2K. This results in moderate magnetic entropy change ( SM) of 7.9 J/kg-K, adiabatic temperature change (ΔTad) of 3.85 K along with RCP value of 146 J/kg for a field change of 70 kOe. The short range magnetic correlation spread over a much higher temperature region than the spin freezing temperature is responsible for the achievement of substantial SM and ΔTad values with asymmetric temperature dependencies. This is one of those few rare-earth based compounds that exhibit considerable magnetocaloric effect despite having low moment rare-earth ion. We believe that these results establish Pr2Ni0.95Si2.95 to be an interesting magnetic material not only for its possible technological application but for fundamental understanding as well.
4