France



The use of a significant share of consumed energy is dedicated to refrigeration and heating systems. Hence, innovating the current state of refrigeration and heating systems to improve their efficiencies is a critical challenge to be tackled, especially under the quest of global warming mitigation. Magnetocaloric (MC) systems are one of the most promising innovations, as their theoretical coefficient of performance (COP) is significantly larger than that of vapor-compression systems. MC systems rely on the reversible magnetocaloric effect where applying (or removing) magnetic fields (H) changes the thermodynamic state of the material in such a way that, in adiabatic processes, the temperature increases (or decreases) [1].&lt;br/&gt;&lt;br/&gt;Although the development of magnetocaloric systems is now in a relatively mature phase, the geometrical and operating optimization is still a time consuming task. One way of optimizing a system is by predicting what are the performance values if using a set of geometrical and operating parameters. In that respect, the recent interest of artificial intelligence lead to the development of several statistical learning techniques and respective algorithm implementations [2]. These techniques have recently been used in a wide number of reserarch fields. One specific class of statistical techniques concerns supervised learning, where regressors (or classifiers) are trained based on datasets with continuous (or bolean) output values [2].&lt;br/&gt;&lt;br/&gt;In this work, different statistical learning regressors are applied, and critically compared, in predicting the three main performance values for a generic magnetocaloric heat pump: no load temperature span, heating power and COP [4]. Four different regressors were used: ordinary least squares (ols), ridge, lasso, and K-nearest neighbors (KNN). The used dataset was obtained with the recently developed heatrapy framework to compute caloric systems [5]. The minimum average relative error of the temperature span, heating power and COP were 23%, 29% and 31%, respectively, as can be seen in Fig. 1. The Lasso regressor shows the best performance with the order of the polynomials. The best order of polynomials range between 3, for the heating power, to 5, for the COP. The worse performance in predicting the three performance values occurs when using the KNN regressor. Furthermore, the application of regressors to the dataset is more adequate to evaluate the temperature span rather than energetic performance values.

**References:**

[1] A. Kitanovski et al., Magnetocaloric Energy Conversion. Springer (2015).&lt;br/&gt;[2] C. M. Bishop, Pattern recognition and machine learning. Springer (2006).&lt;br/&gt;[3] C. Aprea et al., Int. J. Refrig. 82 (2017) 238.&lt;br/&gt;[4] D. J. Silva et al., Energy and AI, 2 (2020) 100030.&lt;br/&gt;[5] D. J. Silva et al., SoftwareX 7 (2018) 373.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/4552fbdd65b2086aba0ca2c6a67e6fc9.jpg)
Figure 1: Minimum average relative error of the temperature span, heating power and COP for the different used regressors.



INTERMAG 2021

 Predicting the performance of magnetocaloric heat pumps using statistical learning algorithms

INTERMAG 2021 Welcome Video



	
<figure class="video_container">
 <iframe src="https://screencast-o-matic.com/player/crflqFVnUpZ?v=6&ff=1&title=0&controls=1" width=640 height=360 frameborder="0" allowfullscreen="true"> </iframe>
</figure>

Want to buy a ticket? 

Register for INTERMAG 2021 here

[![](https://assets.underline.io/uploads/markdown_image/1/image/17ce76e7960db633258024eef8d6da7f.png)](https://intermag.org/register)

Got your ticket, but still can’t access the event?

Contact us and our team will look into it.

The use of a significant share of consumed energy is dedicated to refrigeration and heating systems. Hence, innovating the current state of refrigeration and heating systems to improve their efficiencies is a critical challenge to be tackled, especially under the quest of global warming mitigation. Magnetocaloric (MC) systems are one of the most promising innovations, as their theoretical coefficient of performance (COP) is significantly larger than that of vapor-compression systems. MC systems rely on the reversible magnetocaloric effect where applying (or removing) magnetic fields (H) changes the thermodynamic state of the material in such a way that, in adiabatic processes, the temperature increases (or decreases) [1]. Although the development of magnetocaloric systems is now in a relatively mature phase, the geometrical and operating optimization is still a time consuming task. One way of optimizing a system is by predicting what are the performance values if using a set of geometrical and operating parameters. In that respect, the recent interest of artificial intelligence lead to the development of several statistical learning techniques and respective algorithm implementations [2]. These techniques have recently been used in a wide number of reserarch fields. One specific class of statistical techniques concerns supervised learning, where regressors (or classifiers) are trained based on datasets with continuous (or bolean) output values [2]. In this work, different statistical learning regressors are applied, and critically compared, in predicting the three main performance values for a generic magnetocaloric heat pump: no load temperature span, heating power and COP [4]. Four different regressors were used: ordinary least squares (ols), ridge, lasso, and K-nearest neighbors (KNN). The used dataset was obtained with the recently developed heatrapy framework to compute caloric systems [5]. The minimum average relative error of the temperature span, heating power and COP were 23%, 29% and 31%, respectively, as can be seen in Fig. 1. The Lasso regressor shows the best performance with the order of the polynomials. The best order of polynomials range between 3, for the heating power, to 5, for the COP. The worse performance in predicting the three performance values occurs when using the KNN regressor. Furthermore, the application of regressors to the dataset is more adequate to evaluate the temperature span rather than energetic performance values.

**References:**

[1] A. Kitanovski et al., Magnetocaloric Energy Conversion. Springer (2015). [2] C. M. Bishop, Pattern recognition and machine learning. Springer (2006). [3] C. Aprea et al., Int. J. Refrig. 82 (2017) 238. [4] D. J. Silva et al., Energy and AI, 2 (2020) 100030. [5] D. J. Silva et al., SoftwareX 7 (2018) 373.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/4552fbdd65b2086aba0ca2c6a67e6fc9.jpg)
Figure 1: Minimum average relative error of the temperature span, heating power and COP for the different used regressors.

poster

Perovskite manganites RxA1-xMnO3 (R = Rare earth, A = alkaline earth: 0&lt;x&lt;1) are well-known for their correlated charge, spin and lattice degrees of freedom [1]. While studies on hole-doped manganites are almost complete, the physical properties of electron-doped manganites are still being understood. Recent studies on R0.15Ca0.85MnO3 (R = Y, Gd and Dy) revealed interesting metamagnetic transition in the antiferromagnetically ordered state across which the samples show insulator to metal-like change in the electrical resistivity and also switches from an inverse magnetocaloric effect (MCE) to normal MCE [2]. In this work, magnetic and magnetocaloric properties of electron-doped Pr0.15Ca0.85MnO3 are studied in order to understand the effect of light rare earth substitution at the Ca-site of antiferromagnetic CaMnO3. Polycrystalline Pr0.15Ca0.85MnO3 has been prepared by solid state reaction. The X-ray diffraction (XRD) data were collected at several temperatures (T) from 300 K down to 25 K (Cu Ka, Rigaku). It is found that Pr0.15Ca0.85MnO3 crystallizes in orthorhombic structure (space group Pnma, no. 62, oP20) at 300 K and orders antiferromagnetically at 147 K (TN). Paramagnetic susceptibility follows Curie-Weiss law. The effective paramagnetic moment (meff) and paramagnetic Curie temperature (qp) values are 4.47 mB/f.u and +118 K respectively. The positive qp value indicates the presence of coexisting ferromagnetic interactions in the system. The field dependent magnetization measured at 5 K in fields up to 70 kOe is almost linear suggesting the presence of dominant antiferromagnetic interactions. The magnetization value at 5 K in 70 kOe field is about 0.11 mB/f.u. only. From low temperature XRD, it is observed that the crystal structure changes to monoclinic (space group P21/m, no.11-2, mP20) across the magnetic transition. No metamagnetic transition is evidenced in Pr0.15Ca0.85MnO3 at 5 K in fields up to 70 kOe suggesting robust antiferromagnetic state. Magnetocaloric effect (MCE) has been estimated in terms of isothermal magnetic entropy change (ΔSm) using magnetization vs field (M-H) isotherms measured around TN. Positive MCE (i.e. inverse MCE) is observed below TN [Fig. 1a]. The maximum ΔSm value is about 6.4 Jkg-1K-1 at 137.5 K for 70 kOe field change. The peak in inverse magnetocaloric effect coincides with the magnetostructural transition temperature. The maximum value of isothermal magnetic entropy change, ΔSmmax, at 142.5 K is plotted as a function of magnetic field change (ΔH) and is fitted to a power law, i.e. ΔSmmax µ (ΔH)n [Fig. 1b] and the ‘n’ value is found to be 1.75. It is recalled that a value of n = 2 is expected for antiferromagnets [3] and the decrease in the ‘n’ value could be indicative of the coexistence of ferromagnetic interactions in the system. Though Pr0.15Ca0.85MnO3 shows a strong electrical resistivity value in the antiferromagnetically ordered state, the electrical resistivity value decreased by several order below TN. Non adiabatic polaron hopping conduction is observed above TN and polaron activation energy is found to be ~40 meV. Electron spin resonance studies were performed from 300 K to down to 7 K reveal Dysonian broadening as a result of the intrinsic electrical conductivity of the material. The g-value observed at 300 K is less than the free electron g-value and it increases with decreasing temperature down to TN. Thermopower measurements showed strong temperature dependence near TN and a large negative thermopower of ~40 mV/K was obtained around 100 K.

**References:**

1. Y. Tokura, Rep. Prog. Phys. 69 (2006) 797 2. Lakshman Dhal, Eesha Andharia, N. Shukla, T. Geetha Kumary, A. K. Nigam, S. K. Malik, P. N. Santhosh and R. Nirmala, J. Magn. Magn. Mater. 474 (2019) 215 3. A. Biswas, S. Chandra, T. Samanta, M. H. Phan, I. Das and H. Srikanth, J. Appl. Phys. 113 (2013) 17A902

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/7c4e653992c0eeae931c3f9dc5f0db42.jpg)
Fig. 1 (a) Isothermal magnetic entropy change (ΔSm) vs temperature for various field changes and (b) Maximum magnetic entropy change vs ΔH at TN and the fit to power law, (ΔH)n, with n = 1.75.

 Inverse Magnetocaloric Effect and the Magnetostructural Transition in Pr0.15Ca0.85MnO3 Manganite

Recently, solid-state magnetic refrigeration (MR) technique has great interest of research as a promising technology based on magnetocaloric effect to replace the conventional cooling techniques. Many literature reviewed La0.45Nd0.25Sr0.3MnO3/5%CuO composite shows convincing magnetocaloric properties. The present study investigate the detail magnetocaloric study of La0.45Nd0.25Sr0.3MnO3/ x Wt% (M)O, (x=0.0, 0.25, and 0.50, M = Cu, Co and Ni) nano-composite samples. Pure phase exchange coupled nanocomposites of magnetically soft oxides were prepared via one-pot autocombustion method and heat treated in air at 1100OC for 12 h. The influence of wt% of transition metal oxide to the composite on the structural, morphology, magnetic and magnetocaloric properties have been investigated using X-ray diffraction, SEM and SQUID. XRD pattern of the composite reveals the presence of a distinct pure and single-phase rhombohedral structure of perovskite-based material La0.45Nd0.25Sr0.3MnO3(LNSMO) with space group R-3c and monoclinic structure of metal oxide(MO) with the ICCD numbers (LNSMO: 01-082-1152). No trace of secondary phase was detected within the sensitivity limit of the experiment. Rietveld refinement was performed using fullprof software to calculate prepared sample’s mass ratio, lattice parameter and cell volume. The lattice parameters aren’t much effected due to presence of very small amount of metal oxide and values are matched with published literature. The SEM techniques were used to identify the obtained nano-structural morphologies. The NiO/CoO/CuO material were located at the grain boundaries with uniform homogeneity due to the uniaxial pressure effect. Temperature dependence of field-cooled (FC) curve exhibits second-order transition with Tc near room temperature. The temperature depending of magnetization as a function of applied field for all the samples set were measured and conformed the paramagnetic behavior above the transition temperature and ferromagnetic behavior below the transition temperature. The isothermal magnetic entropy change (ΔSm) was calculated to evaluate the magnetocaloric property of the material. The ΔSm have a maximum value around the vicinity of the transition temperature and decrease with any further increase of the temperature. The maximum value of ΔSm for 5T field were found near room temperature. All the values of ΔSm are higher than ferrites oxides. The reason could be the grain size and particle distribution. The uniform distribution of NiO and CoO particles in the grain boundaries than CuO gives the enhance value of entropy change with CoO and NiO compare to CuO. Also, the grain size of La0.45Nd0.25Sr0.3MnO3 for CoO has higher as compare to CuO. This is the effectiveness of one pot autocombustion technique in achieving effective growth of fine grains of respective phases. Cooling efficiency can be calculated from the peak position of ΔSM called relative cooling power (RCP). RCP values of our nanoparticle samples being the maximum for 5wt%CoO. These results make our compounds promising materials for magnetic refrigeration technology. Our study suggests that the La0.45Nd0.25Sr0.3MnO3/ x Wt% (A)O, (x=0.0, 0.25, and 0.50, A = Cu, Co and Ni) nano-composite materials with various morphologies could exhibit interesting magnetocaloric properties near room temperatures

**References:**

1. Tishin, A. M., & Spichkin, Y. I. (2016). The magnetocaloric effect and its applications. CRC Press. 2. GschneidnerJr, K. A., Pecharsky, V. K., & Tsokol, A. O. (2005). Recent developments in magnetocaloric materials. Reports on progress in physics, 68(6), 1479. 3. Gschneidner Jr, K. A., & Pecharsky, V. K. (2000). Magnetocaloric materials. Annual Review of Materials Science, 30(1), 387-429.. 4. Fkhar, L., Mahmoud, A., Boschini, F., Schrijnemakers, A., Hamedoun, M., Benyoussef, A., ... & Mounkachi, O. (2020). A study of magnetic and magnetocaloric properties of 0.95 (La0. 45Nd0. 25Sr0. 3MnO3)/0.05 CuO composites prepared by spray drying. Inorganic Chemistry Communications, 119, 108129. 5. Fkhar, L., Lamouri, R., Mahmoud, A., Boschini, F., Hamedoun, M., Ez-Zahraouy, H., ... & Mounkachi, O. (2020). Enhanced Magnetic and Magnetocaloric Properties of La0. 45Nd0. 25Sr0. 3MnO3/CuO Composite. Journal of Superconductivity and Novel Magnetism.. 6. Kallel, N., Kallel, S., Hagaza, A., & Oumezzine, M. (2009). Magnetocaloric properties in the Cr-doped La0. 7Sr0. 3MnO3 manganites. Physica B: Condensed Matter, 404(2), 285-288. 7. Kumaresavanji, M., Sousa, C. T., Pires, A., Pereira, A. M., Lopes, A. M. L., & Araujo, J. P. (2014). Room temperature magnetocaloric effect and refrigerant capacitance in La0. 7Sr0. 3MnO3 nanotube arrays. Applied Physics Letters, 105(8), 083110. 8. Fang, Y. K., Yeh, C. C., Chang, C. W., Chang, W. C., Zhu, M. G., & Li, W. (2007). Large low-field magnetocaloric effect in MnCo0. 95Ge1. 14 alloy. Scripta materialia, 57(6), 453-456.

**Images:**

 Magnetocaloric study of La0.45Nd0.25Sr0.3MnO3/ MO (MO=CuO, CoO and Ni) nanocomposites

The precise study of the bonding in materials is an important part of material characterization. The Maximum Entropy Method (MEM) is a powerful tool for deriving accurate electron density distribution in crystalline materials using experimental data. The MEM provides us with the least biased deduction, which is compatible with certain given information. The bonding nature and the distribution of electrons in the bonding region can be clearly visualized using this technique. MEM electron densities are always positive and even with the limited number of data, one can evaluate reliable electron densities resembling true densities. In this present work, the attention is focused on the growth, structural, and electron density distribution of Gd3Fe5-xAlxO12 (0 ≤ x ≤ 1) using X-ray powder data. Gd3Fe5-xAlxO12; x = 0.0, 0.25, 0.50 and 1.0 samples were synthesized via auto combustion method and heat-treated in air at 1100oC for 12 h. The phase identification and lattice parameter of the powders performed using X-ray diffraction data. The Rietveld analysis was performed for the present analysis, which proves the synthesized material of Gd3Fe5-xAlxO12 (0 ≤ x ≤ 1) stays in high-crystalline with single-phase garnet structure, the structural and profile parameter were compared with the reported results from JCPDS-48-0077. The Rietveld analysis was performed using JANA 2000 software package, the results show that the lattice parameters and the cell volume decrease with the increasing the Al, this is clearly evidencing the Al content is replacing the Fe. From this analysis, we observed that a gradual shift in the peaks from the lower angle to the higher angle as compared to the pure Gd3Fe5O12. The lattice parameters for Gd3Fe5-xAlxO12 decreases from 12.50A to 12.44A due to smaller ionic radii of Al3+ ion (0.535A) replacing Fe3+ ions (0.645A). The atomic site occupancy and bond distance analysis for each site were derived from Rietveld refinement for Gd3Fe5-xAlxO12, the Al is preferred to the tetrahedral site for x ≤ 0.5 and prefers both sites; tetrahedral and octahedral thereafter. The structure factors from the Rietveld analysis were further taken to determine the electron density distribution analysis by MEM method. The bonding nature and radii of atoms/ions and interaction between the atoms were clearly analyzed and represented by 3D, 2D, and 1D MEM maps. The mid bond electron densities between atoms were analyzed. The two-dimensional MEM electron density maps on the (001) and (202) planes, and the one-dimensional electron density profiles along Fe-O, Gd-O, and O-O directions were calculated. From this analysis, the substitution of Al3+ ions actually reduces the lattice parameter and volume of the Gd3Fe5-xAlxO12. The 2D map shows the depletion of charge density in the valence regions are clearly seen. Our study suggests that the Al3+ doped Gd3Fe5O12 nano-materials could exhibit interesting electronic properties for data storage applications.

**References:**

1. Lassri, H., Hlil, E. K., Prasad, S., & Krishnan, R. (2011). Magnetic and electronic properties of nanocrystalline Gd3Fe5O12 garnet. Journal of solid state chemistry, 184(12), 3216-3220. 2. Nakamoto, R., Xu, B., Xu, C., Xu, H., & Bellaiche, L. (2017). Properties of rare-earth iron garnets from first principles. Physical Review B, 95(2), 024434. 3. Ali, K. S., Saravanan, R., Israel, S., & Arda, L. (2010). Localized ferromagnetic charge ordering through charge density analysis in nano sized diluted magnetic semiconductor Co2+: ZnO. Physica B: Condensed Matter, 405(7), 1763-1769. 4. Ali, K. S., Saravanan, R., Pashchenko, A. V., & Pashchenko, V. P. (2010). Local distortion in Co-doped LSMO from entropy-maximized charge density distribution. Journal of Alloys and Compounds, 501(2), 307-312. 5. Saravanan, R., & Rani, M. P. (2011). Metal and alloy bonding-an experimental analysis: Charge density in metals and alloys. Springer Science & Business Media.

**Images:**

 Structural and Electronic properties on Gd3Fe5-xAlxO12 (x= 0.25, 0.5, 1.00) using Rietveld, Maximum Entropy Method (MEM)

The family of manganese oxides with the general formula R1-xAxMnO3 (R=lanthanide and A=divalent alkaline earth) has a non-negligible potential in magnetic cooling particularly because of the high chemical stability and interesting magnetocaloric properties at room temperature [1]. In this work, we report a detailed study of the thermodynamic performance of Pr0.6Sr0.4MnO3 in a functional magnetic cooling system up to the theoretical limit of its magnetocaloric effect. In this context, a one-dimensional numerical model of the AMR cycle is developed, while the required magnetocaloric properties of the oxide are calculated in the framework of Mean Field Theory (MFT). The obtained results are deeply analyzed and then compared with the reference Gadolinium (Gd) metal. Moreover, it is revealed that using Pr0.6Sr0.4MnO3 as refrigerant requires high magnetic fields to generate reasonable temperature spans for application. Under a magnetic field of 8 T, the studied oxide can generate an AMR temperature span of 33 K. However, such a span can be achieved by employing Gd and using only a magnetic field source of 1.45 T. The thermodynamic efficiency was evaluated by directly considering the involved magnetic forces between the Pr0.6Sr0.4MnO3 regenerator and the magnetic field source. The associated magnetic work is numerically calculated by using Altair Flux 3D software. Particularly, it was found that the coefficient of performance (COP) can be enhanced by a factor of two when balancing magnetic forces (Fig. 1) despite of the moderate magnetocaloric effect shown by Pr0.6Sr0.4MnO3 under practical fields (1 to 2 T). On the other hand, the hidden thermodynamic performances of this compound are also pointed out by investigating its magnetocaloric properties in the theoretical limit zone. In this way, our calculations unveil that is possible to reach cold source temperatures under -22°C even when using helium as heat transfer fluid (Fig. 2). Finally, a magnetic cooling device working with Pr0.6Sr0.4MnO3 as refrigerant and fully powered by solar cells is proposed.

**References:**

[1] M. Balli, S. Jandl, P. Fournier, and A. Kedous-Lebouc, Appl. Phys. Rev. 4, 021305 (2017).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/c8bd35af86f03d8a8875379d767e6ec9.jpg)
Fig. 1 The simulated magnetic forces in the case of one and two separated regenerators of Pr0.6Sr0.4MnO3.

![](https://assets.underline.io/uploads/markdown_image/1/image/1c021cfc7ac84b35b464f05f9437c523.jpg)
Fig. 2 Temperature evolution of the hot and cold ends as a function of the number of cycles for using helium as transfer fluid

 On the real potential of R1-xAxMnO3 oxides in magnetic cooling: Pr0.6Sr0.4MnO3 as a case of study

The materials that exhibit mutual control of more than one property by external stimuli such as magnetic, electric and stress fields are known as multifunctional material. They are useful to design smart sensors and can be extended as a functional refrigerant for cooling. The magnetic refrigerator is an alternative to the conventional refrigerators as they do not emit toxic greenhouse gases to the environment. However, the product cost of a magnetic refrigerator is expensive due to the raw material and magnetic components required to achieve the maximum entropy change at a given magnetic field strength (≤1.5 T). Assuming the combination of electric and magnetic properties may yield adiabatic temperature change at low (magnetic/electric) field, the piezoelectric/magnetic nanocomposites are proposed. Poly(methyl methacrylate)/Poly(vinylidenefluoride-co-hexafluoropropylene) (PMMA/PVDF-HFP) blend and PMMA/PVDF-HFP/Zn0.5Cu0.5Fe2O4 flexible nanocomposites were fabricated by solution casting method. Pristine polymer blend and PMMA/PVDF-HFP/ferrite nanocomposite are having 67 and 62% of F(b), respectively. Zn0.5Cu0.5Fe2O4 and PMMA/PVDF-HFP/Zn0.5Cu0.5Fe2O4 nanocomposite show superparamagnetic nature with the saturation magnetization of 44 and 5.3 emu/g, respectively. The composite film is thermally stable up to 753 K and later it shows 80% of weight loss. The magnetic entropy change (DSM) of Zn0.5Cu0.5Fe2O4 and PMMA/PVDF-HFP/Zn0.5Cu0.5Fe2O4 show interesting table-like flat variation with respect to magnetic field (0 to 40 kOe) and temperature (100 to 400 K). The DSM (refrigerant capacity) at 300 K is found to be -0.69 (293 J/kg) and -0.094 (40 J/kg) J/kg/K for Zn0.5Cu0.5Fe2O4 and PMMA/PVDF-HFP/Zn0.5Cu0.5Fe2O4, respectively. Ferroelectric polarization change as a function of electric field is analysed at room temperature. The composite film exhibiting both magnetism and ferroelectricity is proved and would be attractive to design different compositions to increase the temperature variation for cooling devices.

**References:**



**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/db2e5a1e5dec581473008e717fb6b89a.jpg)
Figure 1 (A) X-ray diffraction patterns, (B) and (C) M-H hysteresis loops of Zn0.5Cu0.5Fe2O4 and PP/Zn0.5Cu0.5Fe2O4 nanocomposite respectively.

 Multiferroic Electroactive Polymer Blend/Ferrite Nanocomposite Film for Cooling Devices

The current interest in developing solid-state refrigeration devices has encouraged the investigation of caloric effects linked to first-order phase transitions in different families of solids [1]. Among them, (Ni, Mn)-based Heusler-type magnetic shape memory alloys have deserved particular attention due to their multicaloric nature [2]. For properly chosen compositions, martensitic transformation (MT) in Ni-Fe-Ga alloys appears around room temperature (RT) [1]. This, together with their good mechanical properties in comparison with other (Ni,Mn)-based alloys (improved with the precipitation of γ phase [3]), and the facts that with the partial replacement of Fe by Mn in the Ni55Fe19Ga26 alloy γ phase remains in a low concentration, increases the entropy of the MT, not significantly modifying the MT temperature [3], motivated present investigation. Two bulk Ni55Fe11Mn7Ga27 samples were prepared from pure elements (≥ 99.9 %) by arc-melting and spark plasma sintering (SPS) (bulk and SPS samples, respectively), and their phase constitution, as well as magnetocaloric (MC) and elastocaloric (eC) effects studied across the MT. The SPS sample was a dense disc (ρ ≈ 8.4 g.cm-3; ∼ 99 % in respect of theoretical), with diameter and height of 15 and 2.2 mm, sintered from a powdered arc-melted ingot at constant pressure of 30 MPa in a Labox-210 SPS system. The time evolution of vertical displacement and temperature during the process is given in Fig. 1(a). Samples received homogenization annealing at 1273 K for 24 h, followed by a chemical ordering treatment at 773 K (1 h). Start and finish temperatures of the structural transition were determined from the 5 and 95 %, respectively, of the DSC peak area. RT phase constitution was determined by X-ray diffraction (XRD). As Fig. 1(b) shows, the structural transition for both samples is close to RT with similar temperatures. This highlights the effectiveness of SPS technique for consolidating (Ni-Fe-Ga)-based alloys. XRD patterns, shown in Figs. 1(c) and (d), respectively, were obtained with the X-ray beam impinging on a plane perpendicular to the gradient during solidification and on the surface of the sintered disc. In both patterns, a L21-type austenite (AST) is the major phase, while cubic γ phase [4] is also detected (∼ 6 %vol.). However, notice that the peak intensities in the XRD pattern for the SPS sample correspond to an isotropic polycrystal, whereas for the bulk arc melted one AST grains preferentially grow with the [1 0 0] direction oriented along the thermal gradient during solidification. Partial preferential and non-preferential grain orientation in both samples was confirmed by SEM observations (not shown). Magnetization measurements were done by VSM in a PPMS Dynacool system. M(T) curves, under 5 mT and 5 T following zero-field-cooling (ZFC) and field-cooling (FC) regimens (1.0 Kmin-1), are depicted in Fig. 2(a). The magnetic entropy change ΔSM was estimated from the Maxwell relation through the reverse MT from a set of M(T) curves (shown in Fig. 2(b)), measured up to an applied magnetic field μ0Hmax of 5 T. As shown in Fig. 2(c), bulk sample exhibits a moderate │ΔSM│max value of 8.2 (12.9) Jkg-1K-1 under μ0ΔHmax = 2 (5) T (tough │ΔSM│max values measured are the largest reported for alloy studied [3]). The eC effect adiabatic temperature change ΔTadme was determined in a home-made facility under a compression uniaxial load change of 120 MPa for parallelepiped-shaped samples cut from the bulk and SPS samples; uniaxial stress for the bulk sample was applied along thermal gradient during solidification. Fig. 2(d) gives the ΔTadme(T) curves for both samples loading and unloading the uniaxial stress. The bulk sample exhibits a │ΔTadme│max of 2.4 K at a load removal of 120 MPa, whereas this magnitude significantly decreases for the SPS sample. This result agrees with the around 1/3 reduction of the eC effect expected for the direction [1 1 0] [5,6]. Hence, the above-mentioned partial crystallographic texture and isotropic nature of the samples, explain the significant difference observed in the eC effect value. In summary, apart from probing the efficacy of SPS technique for the shaping and consolidation of Ni-Fe-Ga alloys exhibiting similar phase transformation temperatures than those obtained for the bulk sample, the MC and eC effects were assessed for bulk samples of the Heusler alloy Ni55Fe11Mn7Ga27. Whereas the MC effect is moderate in comparison with the reported for (Ni-Mn-X)-based alloys (X= Ga, In, Sn, Sb), our results underline their good mechanical strength and the relevant role of crystallographic texture on the eC effect.

**References:**

[1] Ll. Mañosa, A. Planes, Adv. Mater. 29 (2017) 1603607. [2] Ll. Mañosa, A. Planes, Phil. Trans. R. Soc. A 374 (2016) 20150310. [3] S. Kumar Sarkar, A. Biswas, P.D. Babu, S.D. Kaushik , A. Srivastava, V. Siruguri, M. Krishnan, J. Alloys Compd. 586 (2014) 515–523. [4] T. Omori, N. Kamiya, Y. Sutou, K. Oikawa, R. Kainuma, K. Ishida, Mater. Sci. Engin. A 378 (2004) 403–408. [5] F. Xiao, M. Jin, J. Liu, X. Jin, Acta Mater. 96 (2015) 292–300. [6] F. Hernández-Navarro, J.P. Camarillo-Garcia, C.O. Aguilar-Ortiz, H. Flores-Zúñiga, D. Ríos, J.G. González, P. Álvarez-Alonso, Appl. Phys. Lett. 112 (2018) 164101.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/01c19f0fdf5abb60c13512568d4f30f1.jpg)
Fig. 1. (a) Vertical displacement and temperature versus time along the SPS process. (b) DSC curves for bulk and SPS samples (10 K.min-1). Starting and finishing MT temperatures for the bulk (SPS) sample were MS= 276 (278) K, Mf= 269 (262) K, AS= 277 (279) K, and Af= 286 (290) K. (c) and (d) indexed XRD patterns for bulk and SPS samples, respectively (CuKα).

![](https://assets.underline.io/uploads/markdown_image/1/image/62373c7b9f1fcd916d158bd34a413bff.jpg)
Fig 2. ZFC and FC M(T) curves under 5 mT and 5 T across the structural transition (a), isofield M(T) curves under fields from 1 to 5 T (b), and DSM(T) curves (c) through the reverse MT for the bulk arc melted alloy. (d) ΔTadme(T) for both samples under a load and unload applied stress of 120 MPa. Inset: ΔTadme as a function of time at 297 K for the bulk sample.

 Elastocaloric and magnetocaloric effects through the martensitic transformation in bulk Ni55Fe11Mn7Ga27 alloys produced by arc-melting and spark plasma sintering

Recently so-called shape-memory ferromagnets based on Heusler alloys Ni-Mn-Z (Z = Ga, In, Sn, Sb) began to be considered as a promising materials for magnetic cooling technology. These alloys can exceed compounds with gadolinium in the value of the magnetocaloric effect (MCE) for magnetic field of the same magnitude, because the giant values of the MCE are achieved in them due to the structural transformation accompanying the magnetic transition. The electrical and magnetic properties of the alloy are closely related with the state of its microstructure, which can be changed using various methods of intense plastic deformation (e. g. equal channel angular pressing, intense torsional plastic deformation) or thermobaric treatment (TBT) which is the processing of materials under conditions of high temperature (up to 2000 K) and high quasi-hydrostatic pressure (up to 10 GPa). Ingots of Ni45Mn44In11 alloy were prepared by arc melting in an inert atmosphere and subsequently subjected to annealing at 1100 K for 24 h followed by furnace cooling. The temperature dependence of the electrical resistivity ρ(T) were measured by the standart method. The magnetic properties were measured using an MPMS XL7 (Quantum Design) SQUID magnetometer. The magnetic properties were measured in magnetic fields of up to 70 kOe in the temperature range 150-330 K. Thermobaric treatment has a significant effect on both electrical (changing the type of temperature dependence of electrical resistivity from “semiconductor” to “metallic”) and magnetic properties of the Ni45Mn44In11 alloy. In addition, as a result of TBT, the maximum value of electrical resistivity increases in 1,7 times. Despite this the value of the entropy change ΔS before and after TBT practically was not changed (ΔS ≈ 3.6 J/(kg*K)), but its maximum shifted along the temperature scale from 200 K to 300 K. The results of this work were obtained within the state assignment of Minobrnauki of Russia (theme “Spin” No. AAAA-A18-118020290104-2) and “New functional materials for promising technologies: synthesis, properties, spectroscopy and computer simulation” (No. AAAA-A19-119031890025-9), supported in part by the Government of Russian Federation (Decree No. 211, Contract No. 02.A03.21.0006).

**References:**



**Images:**

 Magnetocaloric effect in the alloy Ni45Mn44In11 subjected to the thermobaric treatment

Heusler alloys are promising candidates for micro-magnetic cooling applications thank their desirable behavior. The possibility of tuning the structural transformation temperature by changing the stoichiometry of compounds is one of the most favorable properties in the presented group of alloys. Modulation of valence electron concentration ratio can be used to produce microwires with the required transformation temperature. Preparation by the Taylor-Ulitovsky technique allows easy and fast production of microwires with protective glass coating. In this work, Heusler alloys in the form of glass-coated microwires with different valence electron concentration ratio (e/a) are analyzed. The effect of e/a tuning on martensitic transformation temperature has been studied using indirect magnetocaloric measurements and their proper analysis. Modified Arrott plot and critical exponent analysis point to the fact that the series of Ni2FeGa microwires belong to the same universality. Hence transformation temperature in this series of alloys has been estimated from transformation temperature vs. e/a dependence using the interpolation method. This research was supported by Slovak Grant Agency VEGA 1/0053/19, Slovak Grant Agency grant number APVV-16-0079.

**References:**

V. Franco and A. Conde, International Journal of Refrigeration, Vol. 33(3), p. 465–473 (2010) A. Kharrat and W. Boujelben, Journal of Low Temperature Physics, Vol. 197, p. 357–378 (2019)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/a857e0e0bd2c19692c000c28b9513ebf.jpg)

 Modified Arrott plot and critical exponent study on series of magnetocaloric Ni2FeGa based Heusler glass-coated microwires

Heusler alloys X2YZ are well-known materials for their various properties which depend on their chemical composition and structure [1]. Their unique features such as shape memory effect, magnetocaloric effect, superconductivity, or spin polarisation have been observed and well described in the form of bulk, rapidly quenched ribbons, glass-coated microwires, or thin films [2]. Therefore, one fundamental question arises: whether the peculiar properties of Heusler alloys will also remain in the nanoscale? There are several technology areas that may be enhanced using Heusler alloys, especially in the field of spintronics, sensors and actuators, biological application based on hypo- or hyperthermia, or drug delivery [3]. The preparation of a large amount of Heusler alloys in the form of nanomaterials with a well-defined shape remains still a challenge. Here, one of the best possible methods consists of the fast and straightforward electrochemical approach [4-5]. With the presented template-assisted deposition using nanoporous anodic alumina membrane, it is possible to prepare a considerable amount (109 nanowires/cm2) of nanodisks or 1D nanowires with uniform morphology and chemical composition. Thanks to the possibility of co-deposition of various metals such as Fe, Co, Ni, Mn, In, Ga, Sn, etc., many types of ternary and quaternary Heusler alloys can be prepared via electrochemical deposition in the various form [6]. We present the possibility of the preparation of Heusler nanowires via electrochemical deposition in nanoporous anodic alumina template with well-defined shape and morphology. Depending on the chemical composition, the prepared Heusler nanowires show different physical properties such as high spin polarisation (Co2FeZ based Heusler alloys) and magnetocaloric or shape memory effect (Ni2FeZ-based). Electrodeposited Co2FeSn nanowires with a diameter of 60 nm and length 10 μm exhibit B2 disordered phase, high curie temperature (&gt;1000 K) and easy magnetization axis parallel to the nanowires axis. Point Contact Andreev Reflection spectroscopy (PCAR) performed with Nb superconductive tip revealed high spin polarisation up to 100 %. The presented approach allows the preparation of promising nano scaled Heusler alloys suitable for applications in spintronics or high-density magnetic data recording [7]. On the other hand, Ni2FeGa-based Heusler alloys are well known as shape memory or magnetocaloric materials. Electrodeposited nanowires with a diameter of 60 nm and length of 7.5 μm show unusual magnetic behavior which points to the structural transformation in the temperature range of 340-395 K. Magnetic nanowires with shape memory effect offer new opportunities in various areas of research such as drug delivery or nanorobotics. Template-assisted electrochemical deposition enables the preparation of a large amount of uniform Heusler nanowires with various chemical composition and related physical properties. Acknowledgments: This research was supported by Slovak Grant Agency VEGA 1/0053/19, Slovak Grant Agency grant number APVV-16-0079.

**References:**

[1] T. Graf, C. Felser, S.S.P. Parkin, Prog. Solid State Chem., 39, 1–50 (2011). [2] J. Alam, C. Bran, et al., J. Magn. Magn. Mater., 513, 167074 (2020). [3] I. Koh, L. Josephson, Sensors, 9, 8130 (2009). [4] L. Galdun, et al., ACS Appl. Nano Mater., 1 (12), 7066–7074 (2018). [5] S. Khan, et al., J. Magn. Magn. Mater., 465, 462−470 (2018). [6] M. Sharma, et al. AIP Advances, 9, 125054 (2019). [7] L. Galdun et al., ACS Appl. Nano Mater., 3, (8), 7438-7445 (2020).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/1ec077b4141b4eb7f7eae8af375ea8f5.jpg)
Figure 1, Representation of free-standing Heusler nanowires whose properties and application possibilities strongly depend on their chemical composition.

 Functional Heusler Nanowires

Materials with a giant magnetocaloric effect near ambient temperature have been a subject of intense investi-gation in the last 20 years due to their high technological potential as magnetic refrigerators or heat pumps, and as thermomagnetic regenerators for thermal energy harvesting1. HVAC (heating, ventilation and air conditioning) technologies based on the magnetocaloric effect could lead to more environmentally friendly systems, due to higher energy efficiency and the lack of greenhouse gases1. These materials are characterized by their strong spin-lattice coupling that manifests itself in first-order phase transitions (FOPT) during which the materials undergo simultaneous or closely coupled magnetic and structural transitions. These transitions may be controlled by extrinsic (temperature, magnetic field, pressure, time) and intrinsic (composition, morphology) parameters. In the referred applications, the devices require relatively high operating frequencies (of the order of 10 Hz) in order to maximize their efficiency. Such a technological requirement implies the utilized materials should complete their FOPT rapidly (for 10Hz operating frequencies, the FOPT timescale should be in the order of few milliseconds), however, unfortunately, there are relatively few studies dedicated to these materials transition kinetics2–8. From the few reports dedicated to this topic, the ones focused on the La-Fe-Si family stand out. In these, it has been shown that the shape and the interactions between particles influence the magnetic hysteresis and the time-scale of the transitions significantly4. In the Gd-Si-Ge family of materials, the Gd5Ge4 composition has been subject of relevant kinetic studies which unveiled its magnetic glass state and a kinetically arrested AFM-FM transition5. To our knowledge there is only one preliminary study for the Gd5Si2Ge2 composition, where it has been shown that the FOPT transition can last up to 300 seconds to be completed - an impressively long time considering the required operating frequencies mentioned above6. Heusler alloys such as Ni-Mn-In exhibit a giant inverse-magnetocaloric effect along their AFM-PM transition, and display typically faster transition kinetics7. Our work will thus fit in an effort to shed light on the kinetics of the FOPTs which occur in these most prominent giant magnetocaloric material families. To accomplish this, we will thoroughly study the kinetics of FOPT, encompassing the magnetic transitions (through time-dependent SQUID measurements) and the (de) coupled structural transitions time-dependencies, their electron transport kinetics as a function of temperature, magnetic field, composition and micro-structure and then attempt to theoretically model the observed phenomena. The growing surge of interest in these unusual transitions is not only due to the pursuit of application-ready materials, but also to attempt to provide answers to purely scientific questions such as how the transition kinetics of the different material families compare and how they change with different magnetovolume couplings and different morphologies (macro, micro and nanostructured). Bulk samples of the above mentioned materials families were produced and their single-phase nature, magnetic and structural properties were assessed via Scanning Electron Microscopy, SQUID magnetometry and X-ray Diffraction. As preliminary work, we have followed the time dependent magnetic field (SQUID) measurement protocol done in8 for a bulk (6x2.5x2 mm) Gd5Si2Ge2 sample. We analyze the evolution of the magnetization of the sample through time, while subjected to initially increasing and then paused fields: the fields are increased with a constant sweep rate of 70 mT/s until a given Hpause value, which is maintained for 300 seconds. Not only is magnetic hysteresis detected, but also a significant difference is observed between magnetization vs time curves obtained by two different measurement protocols: 1) a thermal “reset”, where in between measurements the sample is heated up to 45K above Tc=270K and cooled down again before the following measurement, and 2) no thermal “reset”, where the sample temperature is unchanged before subsequent measurements (see figure 1). In both, one can see once the field is stopped (at t=tpause), the rate of increase of the magnetization slows down, but keeps evolving in time. We made measurements for different temperatures a few degrees above Tc (T=283K in the presented results) so that the sample begins in its paramagnetic state and the transition to the high magnetization state must be induced by the applied field. At the given temperature, the field is increased at 70 mT/s rate until a given field value is reached (HPause) and the M(H) curves show a relaxation behavior - vertical lines in figure 2. As can be seen, the relaxation has non monotonous behaviour, increasing until a maximum at the critical field value around 2 T, and then decreasing again for higher Hpauses. The measurements followed protocol 1, where the relaxation is more significant.

**References:**

1) V. Franco, J. S. Blázquez and J. J. Ipus, Prog. Mater. Sci. Vol. 93 p. 112–232 (2018) 2) V. Basso, M. Piazzi and C. Bennati, Physica Status Solidi B Vol. 255, Issue 2 (2017) 3) A. Biswas, Y. Mudryk and A. K. Pathak, J. Appl. Phys. 126, 243902 (2019) 4) M. G. Zavareh, Y. Skourski and K. P. Skokov, Phys. Rev. Applied 8, 014037 (2017) 5) J. D. Moore, G. K. Perkins and K. Morrison, J. Phys.: Condens. Matter Vol. 20, 465212 (2008) 6) J. Leib, J. E. Snyder and T. A. Lograsso, J. Appl. Phys. 95, 6915 (2004) 7) L. Pfeuffer, T. Gottschall and T. Faske, Phys. Rev. Materials 4, 111401(R) (2020) 8). E. Lovell , A. M. Pereira and A. D. Caplin, Adv. Energy Mater. Vol. 5, Issue 6 (2015)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/33ca740aa8e7cb74167544f6139bf403.jpg)
The evolution of the magnetization for a measurement with protocol 1) (M(t) - thermal "reset") and another with protocol 2) (M(t) - no "reset") and the applied field H(t) increasing at 70mT/s and pausing at Hpause=2.2 T.

![](https://assets.underline.io/uploads/markdown_image/1/image/f853dc79d741d8ca99e0daab54dcdaf2.jpg)
The M vs H curves for different maximum paused fields, Hpause. In the inset, the M vs T curves (for constant field of 0.1 T) are shown. The vertical bar at 283K highlights the temperature at which the M vs H curves were obtained.

Predicting the performance of magnetocaloric heat pumps using statistical learning algorithms

Next from INTERMAG 2021

Inverse Magnetocaloric Effect and the Magnetostructural Transition in Pr0.15Ca0.85MnO3 Manganite

Similar lecture

Magnetocaloric effect in Hf1-xTaxFe2By