France



Solid-state cooling devices based on the “caloric” class of functional materials are considered a promising alternative to conventional vapor compression technologies as they allow complete elimination of high-global warming potential (GWP) refrigerants and have high energy efficiency that is theoretically equivalent to 60% of Carnot efficiency. [1] By definition, multicaloric materials exhibit reversible thermal changes that can be driven concurrently or in sequence by more than one type of external energy sources (magnetic field, electric field, strain/pressure). The use of multiple driving forces can bring about larger thermal changes with smaller field magnitudes over broader operating temperature ranges, while reducing energy losses due to hysteresis in one control parameter by shifting it to another.&lt;br/&gt;&lt;br/&gt;Multicaloric materials under current investigation are fraught with issues related to the inclusion of strategically-limited and toxic elements (e.g., Gd&lt;sub&gt;5&lt;/sub&gt;(GeSi)&lt;sub&gt;4&lt;/sub&gt;, FeRh, MnAs; bold=problematic element) or require complex synthesis and processing to realize an acceptable, or even marginal, functional response. Against this backdrop, the MnTX family of compounds [T = transition metal elements (Fe, Ni, or Co); X= main group p-block element (Si, Ge, or Sn] are poised to overcome these limitations since they are made of earth-abundant, non-toxic elements, are scalable for powder production using low-cost, conventional solid-state processing techniques. Select MnTX alloys demonstrate tunable, room-temperature, first-order magnetic phase transition (i.e., magnetostructural transition from a low temperature TiNiSi-type ferromagnetic orthorhombic phase to a high temperature Ni&lt;sub&gt;2&lt;/sub&gt;In-type paramagnetic hexagonal phase) with promising magnetic fields and pressure-driven caloric responses. Though one can tune the magnetic and caloric properties through careful choice of chemical elements that occupy T- and X-sites or their relative concentrations, large chemical diversity presents formidable combinatorial challenges for experimental and theoretical studies.&lt;br/&gt;&lt;br/&gt;To accelerate the discovery of new MnTX materials with low hysteresis magneto-structural phase transitions near room temperature, we apply a machine learning (ML) approach utilizing data from published literature to construct data-driven ML models as illustrated in figure one. To this end, we start by performing a literature review and constructing a database representative of the MnTX materials that undergo the Pnma to P6&lt;sub&gt;3&lt;/sub&gt;/mmc magneto-structural transition. [2]-[21] We found 87 unique observations that have been experimentally synthesized and whose Pnma to P6&lt;sub&gt;3&lt;/sub&gt;/mmc transition temperature on heating (T&lt;sub&gt;H&lt;/sub&gt;) and P6&lt;sub&gt;3&lt;/sub&gt;/mmc to Pnma transition temperature on cooling (T&lt;sub&gt;C&lt;/sub&gt;) have been characterized. The difference between |T&lt;sub&gt;H&lt;/sub&gt; – T&lt;sub&gt;C&lt;/sub&gt;| is indicative of the hysteresis associated with the phase transition. From these 87 observations, we created 2 different methods of representing the data for machine learning. The first was created using element specific features (e.g., valence electron concentration, electronegativity, and size), where each composition is described as a weighted average of these individual elemental properties. In the second method, we performed density functional theory (DFT) to calculate certain properties (e.g., magnetic moment, bond lengths, and unit cell parameters) of end-point compositions for each compound using the open source electronic structure code quantum espresso.[22]–[24] Each alloy or solid solution was then described as a weighted average of the properties of each end point composition. Models were trained using bootstrapped LASSO, support vector regression (SVR), gaussian process regression and random forest methods implanted in the statistical computing language, R.[25] Of these algorithms SVR was the most accurate on the validation set for both the elemental feature space and the DFT feature space. The perfomance of this model is shown in figure two. When compared to each other the model trained on the DFT generated features out-performed the model trained on the features generated from elemental properties. Promising candidates are recommended for experimental validation and feedback for iterative model improvement.

**References:**

[1] J. Lyubina, “Magnetocaloric materials for energy efficient cooling,” &lt;i&gt;Journal of Physics D: Applied Physics&lt;/i&gt;. 2017, doi: 10.1088/1361-6463/50/5/053002.&lt;br/&gt;[2] A. Biswas &lt;i&gt;et al.&lt;/i&gt;, “Designed materials with the giant magnetocaloric effect near room temperature,” &lt;i&gt;Acta Mater.&lt;/i&gt;, 2019, doi: 10.1016/j.actamat.2019.09.023.&lt;br/&gt;[3] J. H. Chen &lt;i&gt;et al.&lt;/i&gt;, “Tuning martensitic transitions in (MnNiSi)0.65(Fe2Ge)0.35 through heat treatment and hydrostatic pressure,” &lt;i&gt;J. Appl. Phys.&lt;/i&gt;, 2018, doi: 10.1063/1.5051551.&lt;br/&gt;[4] J. Liu &lt;i&gt;et al.&lt;/i&gt;, “Realization of magnetostructural coupling by modifying structural transitions in MnNiSi-CoNiGe system with a wide Curie-temperature window,” &lt;i&gt;Sci. Rep.&lt;/i&gt;, 2016, doi: 10.1038/srep23386.&lt;br/&gt;[5] P. Lloveras &lt;i&gt;et al.&lt;/i&gt;, “Giant reversible barocaloric response of (MnNiSi)1- x(FeCoGe)x (x = 0.39, 0.40, 0.41),” &lt;i&gt;APL Mater.&lt;/i&gt;, 2019, doi: 10.1063/1.5097959.&lt;br/&gt;[6] T. Samanta &lt;i&gt;et al.&lt;/i&gt;, “Effects of hydrostatic pressure on magnetostructural transitions and magnetocaloric properties in (MnNiSi)1-x(FeCoGe)x,” &lt;i&gt;J. Appl. Phys.&lt;/i&gt;, 2015, doi: 10.1063/1.4916339.&lt;br/&gt;[7] F. R. Shen &lt;i&gt;et al.&lt;/i&gt;, “Neutron diffraction study on hydrostatic pressure regulated magnetostructural transition and magnetocaloric effect in MnNi1 - XFexSi1 - yGey alloys,” &lt;i&gt;J. Appl. Phys.&lt;/i&gt;, 2020, doi: 10.1063/5.0003056.&lt;br/&gt;[8] Y. Si &lt;i&gt;et al.&lt;/i&gt;, “Magnetostructural transformation and magnetocaloric effect of Sn-bonded Mn0.66Fe0.34Ni0.66Fe0.34Si0.66Ge0.34 composite,” &lt;i&gt;Sci. Rep.&lt;/i&gt;, 2018, doi: 10.1038/s41598-017-18240-x.&lt;br/&gt;[9] Z. Y. Wei &lt;i&gt;et al.&lt;/i&gt;, “Unprecedentedly Wide Curie-Temperature Windows as Phase-Transition Design Platform for Tunable Magneto-Multifunctional Materials,” &lt;i&gt;Adv. Electron. Mater.&lt;/i&gt;, 2015, doi: 10.1002/aelm.201500076.&lt;br/&gt;[10] C. L. Zhang &lt;i&gt;et al.&lt;/i&gt;, “Tunable magnetostructural coupling and large magnetocaloric effect in Mn1−xNi1−xFe2xSi1−xGax,” &lt;i&gt;J. Magn. Magn. Mater.&lt;/i&gt;, 2017, doi: 10.1016/j.jmmm.2017.02.046.&lt;br/&gt;[11] C. L. Zhang &lt;i&gt;et al.&lt;/i&gt;, “Magnetostructural transition and magnetocaloric effect in MnNiSi-Fe2Ge system,” &lt;i&gt;Appl. Phys. Lett.&lt;/i&gt;, 2015, doi: 10.1063/1.4936610.&lt;br/&gt;[12] C. L. Zhang &lt;i&gt;et al.&lt;/i&gt;, “The tunable magnetostructural transition in MnNiSi-FeNiGe system,” &lt;i&gt;Appl. Phys. Lett.&lt;/i&gt;, 2013, doi: 10.1063/1.4823510.&lt;br/&gt;[13] J. Q. Zhao &lt;i&gt;et al.&lt;/i&gt;, “Tunable magnetostructural phase transition and magnetocaloric effect in Mn1−xNi1−xCo2xSi1−xGexsystem,” &lt;i&gt;J. Alloys Compd.&lt;/i&gt;, 2017, doi: 10.1016/j.jallcom.2016.12.156.&lt;br/&gt;[14] K. Deepak and R. V. Ramanujan, “Magnetocaloric properties of low-cost fe and sn substituted mnnisi-based alloys exhibiting a magnetostructural transition near room temperature,” &lt;i&gt;IEEE Trans. Magn.&lt;/i&gt;, 2018, doi: 10.1109/TMAG.2018.2832090.&lt;br/&gt;[15] K. Deepak and R. V. Ramanujan, “Near room temperature giant magnetocaloric effect in (MnNiSi)1-x(Fe2Ge)x alloys,” &lt;i&gt;J. Alloys Compd.&lt;/i&gt;, 2018, doi: 10.1016/j.jallcom.2018.02.018.&lt;br/&gt;[16] P. Dutta, S. Pramanick, S. Chattopadhyay, D. Das, and S. Chatterjee, “Observation of colossal magnetocaloric effect and its dependence on applied hydrostatic pressure in thermally cycled Mn0.53Fe0.47NiSi0.53Ge0.47 alloy,” &lt;i&gt;J. Alloys Compd.&lt;/i&gt;, 2018, doi: 10.1016/j.jallcom.2017.11.379.&lt;br/&gt;[17] S. Ghosh, A. Ghosh, P. Sen, and K. Mandal, “Giant Room-Temperature Magnetocaloric Effect across the Magnetostructural Transition in (MnNiSi)1-x (FeCoGa)x Alloys,” &lt;i&gt;Phys. Rev. Appl.&lt;/i&gt;, 2020, doi: 10.1103/PhysRevApplied.14.014016.&lt;br/&gt;[18] S. Ghosh, P. Sen, and K. Mandal, “Magnetostructural transition and large magnetocaloric effect in (Mn0.6Fe0.4)NiSi1−xAlx (x = 0.06–0.08) alloys,” &lt;i&gt;J. Magn. Magn. Mater.&lt;/i&gt;, 2020, doi: 10.1016/j.jmmm.2019.166345.&lt;br/&gt;[19] D. Kasimov, J. Liu, Y. Gong, G. Xu, F. Xu, and G. Lu, “Realization of magnetostructural coupling in a high temperature region in Mn0.85Co0.3Ni0.85Si1-xGax system,” &lt;i&gt;J. Alloys Compd.&lt;/i&gt;, 2018, doi: 10.1016/j.jallcom.2017.10.295.&lt;br/&gt;[20] L. Lei &lt;i&gt;et al.&lt;/i&gt;, “The magnetostructural transition and magnetocaloric properties in Fe0.6Mn0.4NiSi1- xAl xalloys,” &lt;i&gt;J. Appl. Phys.&lt;/i&gt;, 2020, doi: 10.1063/5.0003261.&lt;br/&gt;[21] Y. Li &lt;i&gt;et al.&lt;/i&gt;, “Windows open for highly tunable magnetostructural phase transitions,” &lt;i&gt;APL Mater.&lt;/i&gt;, 2016, doi: 10.1063/1.4955214.&lt;br/&gt;[22] P. Giannozzi &lt;i&gt;et al.&lt;/i&gt;, “Advanced capabilities for materials modelling with Quantum ESPRESSO,” &lt;i&gt;J. Phys. Condens. Matter&lt;/i&gt;, 2017, doi: 10.1088/1361-648X/aa8f79.&lt;br/&gt;[23] P. Giannozzi &lt;i&gt;et al.&lt;/i&gt;, “Quantum ESPRESSO toward the exascale,” &lt;i&gt;J. Chem. Phys.&lt;/i&gt;, 2020, doi: 10.1063/5.0005082.&lt;br/&gt;[24] P. Giannozzi &lt;i&gt;et al.&lt;/i&gt;, “QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials,” &lt;i&gt;J. Phys. Condens. Matter&lt;/i&gt;, 2009, doi: 10.1088/0953-8984/21/39/395502.&lt;br/&gt;[25] R. R Development Core Team, &lt;i&gt;R: A Language and Environment for Statistical Computing&lt;/i&gt;. 2011.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/5a18e7f6e1ab25eed1a00d6e7a5c447b.jpg)
Figure 1: DFT model performance, phase transition temperature predicted is plotted against the observed experimental data.

![](https://assets.underline.io/uploads/markdown_image/1/image/b9a8448a41632ac63daf60f312ad0c32.jpg)
Figure 2: Workflow of machine learning optimization of MTX alloys.

INTERMAG 2021

 Machine Learning Strategies for Screening Mn-T-X (T=Co, Fe, Ni; X=Si, Ge, Sn) Multicaloric Materials

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.

Solid-state cooling devices based on the “caloric” class of functional materials are considered a promising alternative to conventional vapor compression technologies as they allow complete elimination of high-global warming potential (GWP) refrigerants and have high energy efficiency that is theoretically equivalent to 60% of Carnot efficiency. [1] By definition, multicaloric materials exhibit reversible thermal changes that can be driven concurrently or in sequence by more than one type of external energy sources (magnetic field, electric field, strain/pressure). The use of multiple driving forces can bring about larger thermal changes with smaller field magnitudes over broader operating temperature ranges, while reducing energy losses due to hysteresis in one control parameter by shifting it to another. Multicaloric materials under current investigation are fraught with issues related to the inclusion of strategically-limited and toxic elements (e.g., Gd5(GeSi)4, FeRh, MnAs; bold=problematic element) or require complex synthesis and processing to realize an acceptable, or even marginal, functional response. Against this backdrop, the MnTX family of compounds [T = transition metal elements (Fe, Ni, or Co); X= main group p-block element (Si, Ge, or Sn] are poised to overcome these limitations since they are made of earth-abundant, non-toxic elements, are scalable for powder production using low-cost, conventional solid-state processing techniques. Select MnTX alloys demonstrate tunable, room-temperature, first-order magnetic phase transition (i.e., magnetostructural transition from a low temperature TiNiSi-type ferromagnetic orthorhombic phase to a high temperature Ni2In-type paramagnetic hexagonal phase) with promising magnetic fields and pressure-driven caloric responses. Though one can tune the magnetic and caloric properties through careful choice of chemical elements that occupy T- and X-sites or their relative concentrations, large chemical diversity presents formidable combinatorial challenges for experimental and theoretical studies. To accelerate the discovery of new MnTX materials with low hysteresis magneto-structural phase transitions near room temperature, we apply a machine learning (ML) approach utilizing data from published literature to construct data-driven ML models as illustrated in figure one. To this end, we start by performing a literature review and constructing a database representative of the MnTX materials that undergo the Pnma to P63/mmc magneto-structural transition. [2]-[21] We found 87 unique observations that have been experimentally synthesized and whose Pnma to P63/mmc transition temperature on heating (TH) and P63/mmc to Pnma transition temperature on cooling (TC) have been characterized. The difference between |TH – TC| is indicative of the hysteresis associated with the phase transition. From these 87 observations, we created 2 different methods of representing the data for machine learning. The first was created using element specific features (e.g., valence electron concentration, electronegativity, and size), where each composition is described as a weighted average of these individual elemental properties. In the second method, we performed density functional theory (DFT) to calculate certain properties (e.g., magnetic moment, bond lengths, and unit cell parameters) of end-point compositions for each compound using the open source electronic structure code quantum espresso.[22]–[24] Each alloy or solid solution was then described as a weighted average of the properties of each end point composition. Models were trained using bootstrapped LASSO, support vector regression (SVR), gaussian process regression and random forest methods implanted in the statistical computing language, R.[25] Of these algorithms SVR was the most accurate on the validation set for both the elemental feature space and the DFT feature space. The perfomance of this model is shown in figure two. When compared to each other the model trained on the DFT generated features out-performed the model trained on the features generated from elemental properties. Promising candidates are recommended for experimental validation and feedback for iterative model improvement.

**References:**

[1] J. Lyubina, “Magnetocaloric materials for energy efficient cooling,” Journal of Physics D: Applied Physics. 2017, doi: 10.1088/1361-6463/50/5/053002. [2] A. Biswas et al., “Designed materials with the giant magnetocaloric effect near room temperature,” Acta Mater., 2019, doi: 10.1016/j.actamat.2019.09.023. [3] J. H. Chen et al., “Tuning martensitic transitions in (MnNiSi)0.65(Fe2Ge)0.35 through heat treatment and hydrostatic pressure,” J. Appl. Phys., 2018, doi: 10.1063/1.5051551. [4] J. Liu et al., “Realization of magnetostructural coupling by modifying structural transitions in MnNiSi-CoNiGe system with a wide Curie-temperature window,” Sci. Rep., 2016, doi: 10.1038/srep23386. [5] P. Lloveras et al., “Giant reversible barocaloric response of (MnNiSi)1- x(FeCoGe)x (x = 0.39, 0.40, 0.41),” APL Mater., 2019, doi: 10.1063/1.5097959. [6] T. Samanta et al., “Effects of hydrostatic pressure on magnetostructural transitions and magnetocaloric properties in (MnNiSi)1-x(FeCoGe)x,” J. Appl. Phys., 2015, doi: 10.1063/1.4916339. [7] F. R. Shen et al., “Neutron diffraction study on hydrostatic pressure regulated magnetostructural transition and magnetocaloric effect in MnNi1 - XFexSi1 - yGey alloys,” J. Appl. Phys., 2020, doi: 10.1063/5.0003056. [8] Y. Si et al., “Magnetostructural transformation and magnetocaloric effect of Sn-bonded Mn0.66Fe0.34Ni0.66Fe0.34Si0.66Ge0.34 composite,” Sci. Rep., 2018, doi: 10.1038/s41598-017-18240-x. [9] Z. Y. Wei et al., “Unprecedentedly Wide Curie-Temperature Windows as Phase-Transition Design Platform for Tunable Magneto-Multifunctional Materials,” Adv. Electron. Mater., 2015, doi: 10.1002/aelm.201500076. [10] C. L. Zhang et al., “Tunable magnetostructural coupling and large magnetocaloric effect in Mn1−xNi1−xFe2xSi1−xGax,” J. Magn. Magn. Mater., 2017, doi: 10.1016/j.jmmm.2017.02.046. [11] C. L. Zhang et al., “Magnetostructural transition and magnetocaloric effect in MnNiSi-Fe2Ge system,” Appl. Phys. Lett., 2015, doi: 10.1063/1.4936610. [12] C. L. Zhang et al., “The tunable magnetostructural transition in MnNiSi-FeNiGe system,” Appl. Phys. Lett., 2013, doi: 10.1063/1.4823510. [13] J. Q. Zhao et al., “Tunable magnetostructural phase transition and magnetocaloric effect in Mn1−xNi1−xCo2xSi1−xGexsystem,” J. Alloys Compd., 2017, doi: 10.1016/j.jallcom.2016.12.156. [14] K. Deepak and R. V. Ramanujan, “Magnetocaloric properties of low-cost fe and sn substituted mnnisi-based alloys exhibiting a magnetostructural transition near room temperature,” IEEE Trans. Magn., 2018, doi: 10.1109/TMAG.2018.2832090. [15] K. Deepak and R. V. Ramanujan, “Near room temperature giant magnetocaloric effect in (MnNiSi)1-x(Fe2Ge)x alloys,” J. Alloys Compd., 2018, doi: 10.1016/j.jallcom.2018.02.018. [16] P. Dutta, S. Pramanick, S. Chattopadhyay, D. Das, and S. Chatterjee, “Observation of colossal magnetocaloric effect and its dependence on applied hydrostatic pressure in thermally cycled Mn0.53Fe0.47NiSi0.53Ge0.47 alloy,” J. Alloys Compd., 2018, doi: 10.1016/j.jallcom.2017.11.379. [17] S. Ghosh, A. Ghosh, P. Sen, and K. Mandal, “Giant Room-Temperature Magnetocaloric Effect across the Magnetostructural Transition in (MnNiSi)1-x (FeCoGa)x Alloys,” Phys. Rev. Appl., 2020, doi: 10.1103/PhysRevApplied.14.014016. [18] S. Ghosh, P. Sen, and K. Mandal, “Magnetostructural transition and large magnetocaloric effect in (Mn0.6Fe0.4)NiSi1−xAlx (x = 0.06–0.08) alloys,” J. Magn. Magn. Mater., 2020, doi: 10.1016/j.jmmm.2019.166345. [19] D. Kasimov, J. Liu, Y. Gong, G. Xu, F. Xu, and G. Lu, “Realization of magnetostructural coupling in a high temperature region in Mn0.85Co0.3Ni0.85Si1-xGax system,” J. Alloys Compd., 2018, doi: 10.1016/j.jallcom.2017.10.295. [20] L. Lei et al., “The magnetostructural transition and magnetocaloric properties in Fe0.6Mn0.4NiSi1- xAl xalloys,” J. Appl. Phys., 2020, doi: 10.1063/5.0003261. [21] Y. Li et al., “Windows open for highly tunable magnetostructural phase transitions,” APL Mater., 2016, doi: 10.1063/1.4955214. [22] P. Giannozzi et al., “Advanced capabilities for materials modelling with Quantum ESPRESSO,” J. Phys. Condens. Matter, 2017, doi: 10.1088/1361-648X/aa8f79. [23] P. Giannozzi et al., “Quantum ESPRESSO toward the exascale,” J. Chem. Phys., 2020, doi: 10.1063/5.0005082. [24] P. Giannozzi et al., “QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials,” J. Phys. Condens. Matter, 2009, doi: 10.1088/0953-8984/21/39/395502. [25] R. R Development Core Team, R: A Language and Environment for Statistical Computing. 2011.

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/5a18e7f6e1ab25eed1a00d6e7a5c447b.jpg)
Figure 1: DFT model performance, phase transition temperature predicted is plotted against the observed experimental data.

![](https://assets.underline.io/uploads/markdown_image/1/image/b9a8448a41632ac63daf60f312ad0c32.jpg)
Figure 2: Workflow of machine learning optimization of MTX alloys.

technical paper

Hard magnetic nanocomposites, consisting of a fine mixture between a large magneto crystalline anisotropy phase and a large magnetization phase, are attractive materials both for integration of hardness-tuned magnets in microsystems and for building next-generation high performance permanent magnets for energy conversion technologies. Instead of relying on the hypothetical discovery of new hard magnetic compound, this approach is based on the mastering of the microstructure in a mixture of already known magnetic phases. Theoretical descriptions predicted a potential energy product of 1 MJ/m3, which is twice as large as the one of best Nd2Fe14B magnets produced today [1]. However, exploiting their full potential has appeared to be very challenging since the first reports on exchange spring structures in the 90’s. Indeed, calculations pointed out the absolute necessity to confine the softer phase in nano-sized grains, smaller than twice the domain wall width of the hard phase. In parallel, to maximize the magnetization, the hard phase volume fraction should be restricted to the minimum needed to develop enough coercivity. These requirements cannot be obtained with conventional process for mass scale material fabrication. In this context, investigations of model systems in films can give insights about the interplay between the nanoscale microstructure and the magnetic properties [2]. In particular, the degree of interdiffusion at the soft-hard interface is expected to significantly impact the magnetization reversal but is difficult to address experimentally. Here we focus on the impact of a graded soft-hard interface on magnetic properties. To do so, we recently prepared Co@FePt transition metal-based nanocomposite (NC) films from mass-selected low energy cluster beam deposition technique (MS-LECBD) of Co clusters, in situ embedded in FePt matrix independently produced by alternative electron gun evaporation on the same substrate [3,4,5]. The Co cluster inclusions can be selected in size prior to their deposition, between 1 and 10 nm. The cluster to matrix volume ratio is adjusted controlling the thickness of each Co, Fe, Pt individual layers. The as-prepared Co clusters have been observed by transmission electron microscopy (TEM) and found to crystallise in magnetically soft fcc phase. In this study, we focused on 8 nm Co clusters. The Co to FePt volume fraction was varied from 0 to 50%. Post-deposition annealing of the whole stack is a crucial step to drive the initial Fe and Pt multilayers to the high-magnetic anisotropy L10 phase. However, this inevitably comes with partial intermixing at the interface between the clusters and the matrix. We systematically compared the nanoclusters-based film with same stoichiometry and alloyed films (NF), obtained by sequential evaporation of Co, Fe, Pt thin and continuous layers. Local atomic structures have been thoroughly investigated using polarization-dependent hard x-ray absorption and element-specific spectroscopies (EXAFS, XLD). Transmission electron microscopy (TEM) equipped with energy-dispersive x-ray spectroscopy (EDX) was used to explore the local chemical profile of the composite films (figure 1). The macroscopic magnetic properties were measured in a Superconducting quantum interference device (SQUID) magnetometer together with local and element specific characterisations using circular dichroism (XMCD). We demonstrate the persistence of soft and hard regions and rule out the fully diluted hypothesis. Magnetic measurements (figure 2) showed a very different behaviour upon annealing between NF and NC films. Using a combinatorial approach, we investigated the coercive magnetic field decay when increasing the Co content. The magnetic coercivity is found to be systematically higher in the NC samples, over the whole range of studied composition. EXAFS spectra fitting and numerical modelling of the diffusion mechanism show that magnetically soft Co-rich inclusions remain after annealing, even at the high temperature required to obtain the L10 phase of the matrix. In the case of multilayer nanofilms with Co thin layers, the diffusion leads to homogeneous alloying. Local XMCD revealed the evolution of spin and orbital moments of Fe and Co as a function of the composition, further distinguishing the NC system with respect to fully homogeneous alloy. Studying these materials in model systems synthesised by nanofabrication routes provides interesting insights into the interplay between the microstructure and the magnetic performances in view to develop mass-scale production of such materials. This work is supported by the ANR collaborative project "SHAMAN": Soft in Hard MAgnetic Nanocomposites (2017-2020).

**References:**

1. Skomski R. and Coey J. M. D. Phys. Rev. B 48, 15812 (1993). 2. Balasubramanian B. et al, Scientific reports 4, 6265 (2014) 3. Tournus F. et al, Journal of Magnetism and Magnetic Materials, vol. 323, p. 1868 (2011) 4. Liu Y. et al, Appl. Phys. Lett. 99, 172504 (2011) 5. Paleo C. et al, Interplay between local structure and magnetic properties of graded exchange-coupled Co@FePt nanocomposite films. Phys. Rev. B 7 (2020).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/854d17b4235dd33402f5177028c56401.jpg)
Energy-dispersive x-ray images of a FePt/Co cluster/FePt sample on a Si3N4 substrate, along with the scanning transmission electron microscopy-annular dark-field (STEM-ADF) image.

![](https://assets.underline.io/uploads/markdown_image/1/image/8ac33347c0e0ae0e397c52a2e4bb7528.jpg)
Hysteresis loops at 300 K for two samples with the same Co content and a L10-FePt sample without Co. The NC sample has higher saturation magnetization than the bare L10-FePt ref., which is brought by the Co but retains a sizeable coercivity, whereas the NF is magnetically soft. Inset: Dependence of the coercive field μ0Hc on the Co content for the NF and NC.

 Fcc-Co Clusters in L10-FePt Matrix as Graded-Interface Magnetic Nanocomposite

Low-temperature phase (LTP) MnBi has the hexagonal NiAs structure and retains its ferromagnetic phase up to 613 K [1]. Above 613 K, some Mn atoms from the octahedral sites diffuse into the trigonal-bipyramidal sites, resulting in the antiferromagnetic coupling between Mn at the octahedral and Mn at the trigonal-bipyramidal interstitial sites. Therefore, the magnetic moment sharply drops before reaching the Curie temperature (Tc) of 720 K. Spin reorientation of LTP MnBi occurs at about 90 K from ab-plane (0 - 90 K) to c-axis (T &gt; 90 K) [2]. The spin orientation and magnetocrystalline anisotropy of LTP MnBi can be tuned by partially replacing Bi of MnBi with the third element. Sakuma et al. have performed first-principles calculations on Sn-doped MnBi (MnBi1-xSnx) and found that the Ku increases to about 3 MJ/m3 at x = 0.1 from - 0.5 MJ/m3 at x = 0 and then remains unchanged up to x = 0.3, and TC is 580 K for x = 0 [3]. These calculated Ku and TC for x = 0 are much larger and lower than experimental -0.2 MJ/m3 and 720 K, respectively. Sakuma et al. did not relax MnBi1-xSnx (x &gt; 0) to estimate lattice constants, but used the lattice constants (x = 0) for MnBi1-xSnx (x &gt; 0). Therefore, the calculated Ku and magnetic moment for x &gt; 0 would be inaccurate. In this study, hexagonal crystal structures of LTP MnBi and Sn-doped MnBi in Fig. 1 were used. The LTP MnBi unit cell has two Mn atoms at the octahedral 2a sites of (0, 0, 0) and (0, 0, 1/2) and two Bi atoms at the 2c sites of (1/3, 2/3, 1/4) and (2/3, 1/3, 3/4) [4]. We have relaxed MnBi1-xSnx (x = 0, 0.5) unit cell to obtain lattice constants a and c. Doped Sn decreases the a and c from 4.287 and 6.118 Å for x = 0 to 3.954 and 5.456 Å for x =0.5, respectively. Accordingly, the volume and c/a ratio also decrease from 97.37 Å3 and 1.43 (x = 0) to 73.87 Å3 and 1.38 (x = 0.5), respectively. These relaxed lattice constants (x = 0) are in good agreement with the experimental lattice constants [5, 6]. The WIEN2k package based on density functional theory (DFT) and using the full-potential linearized augmented plane wave (FPLAPW) method was used to conduct the first-principles calculations [7]. In order to calculate spin-polarization and spin-orbit coupling, we used the DFT within the local-spin-density approximation (LSDA) and 19 × 19 × 27 k-point mesh generating 1400 k-points in the irreducible part of the Brillouin zone. Figure 2 shows density of states for MnBi1-xSnx (x = 0, 0.5). The net magnetic moments per formula unit is 3.613 μB/f.u. for x = 0 and 2.517 μB/f.u. for x = 0.5. The net magnetic moment decreases as the Sn concentration increases. The corresponding saturation magnetization in the unit of Tesla (T) monotonically decreases to 0.794 from 0.865 T as the Sn concentration increases to x = 0.5. Calculated magnetic moment of 3.613 μB/f.u., Ku of -0.2 MJ/m3 and TC of 711 Ku for x = 0 are in good agreement with the previously reported results [8]. We will present Ku and magnetic moment as a function of Sn concentration and discuss the discrepancy between Sakuma’s results and this work.

**References:**

[1] A. F. Andersen, Acta Chemica Scandinavica 21, 1543 (1967). [2] Michael A. McGuire, Huibo Cao, Bryan C. Chakoumakos, and Brian C. Sales, Phys. Rev. B 90, 174425 (2014). [3] Akimasa Sakuma, Yuki Manabe, and Yohei Kota, Journal of the Physical Society of Japan 82, 073704 (2013). [4] http://materials.springer.com/isp/crystallographic/docs/sd_1823960, 2020 (accessed 11 November 2020). [5] B. W. Roberts, Phys. Rev. 104, 607 (1956) [6] Wenyong Zhang et al., APL Mater. 7, 121111 (2019) [7] P. Blaha, et al., WIEN2K: An Augmented Plane Wave + Local Orbitals Program for Calculating Crystal Properties (Karlheinz Schwarz, Techn. Universitat, Wien, Austria, 2001) [8] J. Köhler et al., J. Phys.: Condens. Matter 8, 8681(1996)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/153d31d85f39c335487004e072bb1ff7.jpg)
Fig. 1. Crystal structure for (a) MnBi and (b) MnBi0.5Sn0.5

![](https://assets.underline.io/uploads/markdown_image/1/image/1e46ae3b4980f8fe52a17d3b94569416.jpg)
Fig. 2. Density of states for (a) MnBi and (b) MnBi0.5Sn0.5

 Tuning Magnetic Anisotropy of MnBi Permanent Magnet with Sn*

[Introduction] Hard magnetic materials with high Ku, low Ms, and low α have attracted much attention to MRAM application. MnAl alloy with L10 (Pearson symbol: tP4, prototype: CuAu) phase and Mn3Ge alloy with D022 (Pearson symbol: tI8, prototype: Al3Ti) phase have (Ku, Ms, α) = (1.5×107 erg/cm3 [1], 500 emu/cm3 [2], 0.006-0.0175 [3,4]) and (1.2×107 erg/cm3, 100 emu/cm3 [5], 0.03 [6]), respectively. However, these phases are metastable and crystallographic phases other than L10 or D022 may be involved in Mn-Al and Mn-Ge films [3,5]. The formation of ordered phases is delicately influenced by the film composition. In the present study, Mn-Al and Mn-Ge films are prepared by using a molecular beam epitaxy system with a reflection high-energy electron diffraction (RHEED) facility, which can reveal the crystallographic property during film formation. The effects of composition on the structural and magnetic properties are investigated. [Experimental Procedure] 20-nm-thick MnxAl100-x and MnyGe100-y (at. %) films were formed on Cr(001) single-crystal underlayers at 300 °C. The Mn contents, x and y, were varied from 40 to 100 at. %. The crystal structure and the crystallographic orientation relationship between film and underlayer were determined by RHEED. The order degrees of L10 and D022 phases were estimated by XRD. The surface morphology was observed by AFM. The magnetization curves were measured by VSM. [Results and Discussion] Fig. 1(a) shows the RHEED patterns observed for Mn-Al films with different compositions. Diffraction patterns from L10(001) and A12(001) surfaces [Fig. 1(b-1)] are overlapped for the films with x ≤ 50, which involves the L10-MnAl stoichiometry (x = 50), as shown for example in Fig. 1(a-1). Although metastable L10 phase is formed, A12 (Pearson symbol: cI58, prototype: α-Mn) phase is coexisting with L10 phase in the films. Diffraction pattern from only L10(001) surface [Fig. 1(b-2)] is observed for the films with Mn-rich compositions of x = 58-72, as shown in Fig. 1(a-2). The result shows that L10(001) single-crystal films with the c-axis perpendicular to the substrate surface are successfully obtained. The formation of L10 ordered phase is promoted by using Mn-rich compositions. However, the order degree determined by XRD (not shown here) decreases with increasing the Mn content. With further increasing the Mn content (x = 80-100), diffraction pattern from only A12(001) surface is recognized. Fig. 1(c) summarizes the compositional dependence on crystal structure. Fig. 1(d) shows the crystallographic orientation relationships of L10(001) and A12(001) crystals with respect to Cr(001) underlayer, which are also determined by RHEED. Fig. 1(e) shows the RHEED patterns observed for Mn-Ge films. Diffraction pattern from D022(001) surface [Fig. 1(f-2)] is obtained for the films with y = 68-75 including the D022-Mn3Ge stoichiometry (y = 75). Metastable D022(001) single-crystal films are epitaxially grown on the Cr underlayers. For the films with Mn-rich compositions of y = 83-100, diffraction pattern from A12(001) surface [Fig. 1(f-3)] is recognized, similar to the case of Mn-Al films. For the films with Ge-rich compositions of y = 43-60, diffraction pattern from (001) single-crystal surface with simple cubic structure (Pearson symbol: cP, lattice parameter: a = 0.85 nm) [Fig. 1(f-1)], which does not exist in the bulk Mn-Ge phase diagram, is appearing. In order to determined the crystal structure, characterization by neutron diffraction is necessary. The crystal structures and the epitaxial orientation relationships of Mn-Ge films are summarized in Figs. 1(g) and (h), respectively. Figs. 2(a) and (b) show the compositional dependences on surface roughness values (Ra) of Mn-Al and Mn-Ge films. Flat surfaces with Ra ≤ 0.3 nm are realized for the Mn-Al films with x = 58-72 and the Mn-Ge films with y = 68-75 which are respectively composed of L10 and D022 phases. Fig. 2(c) shows the magnetization curves mesured for Mn-Al films. The Mn58Al42 film consisting of L10 phase shows strong perpendicular magnetic anisotropy [Fig. 2(c-2)]. The magnetization does not saturate at 10 kOe. The magnetic property is apparently reflecting the magnetocrystalline anisotropy of L10 phase. Isotropic magnetization curves are observed for the Mn65Al35 film [Fig. 2(c-3)], though the Mn65Al35 film is consisting of only L10 phase. The reason is due to an influence of low order degree. The Mn50Al50 film also shows isotropic magnetic property due to an influence of coexistence of L10 and A12 phases. Fig. 2(d) shows the magnetization curves of Mn-Ge films composed of only D022 phase. These films do not show perpendicular magnetic anisotropies, since the order degrees are not so high. In the present study, L10-MnAl and D022-Mn3Ge films are prepared under similar experimental conditions. Although flat surfaces which are required for practical applications are realized for both MnAl and Mn3Ge films, strong perpendicular anisotropy is observed only for a MnAl film. The present study shows that it is important achieving a high order degree in addition to obtaining a single phase of L10 or D022 in order to enhance strong perpendicular magnetic anisotropy.

**References:**

[1] A. Sakuma: J. Phys. Soc. Jpn. 63, 1422 (1994). [2] J. H. Park, Y. K. Hong, S. Bae, J. J. Lee, J. Jalli, G. S. Abo, N. Neveu, S. G. Kim, C. J. Choi, J. G. Lee: J. Appl. Phys. 107, 09A731 (2010). [3] M. Hosoda, M. Oogane, M. Kubota, T. Kubota, H. Saruyama, S. Iihama, H. Naganuma, Y. Ando: J. Appl. Phys. 111, 07A324 (2012). [4] S. W. Mao, J. Lu, L. Yang, X. Z. Ruan, H. L. Wang, D. H. Wei, Y. B. Xu, J. H. Zhao: Chn. Phys. Lett. 37, 058501 (2020). [5] A. Sugihara, S. Mizukami, Y. Yamada, K. Koike, T. Miyazaki: Appl. Phys. Lett. 104, 132404 (2014). [6] S. Mizukami, A. Sugihara, S. Iihama, Y. Sasaki, K. Z. Suzuki, T. Miyazaki: Appl. Phys. Lett. 108, 012404 (2016).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/fbfc920f2d2586f3826ea2aaa1ae9088.jpg)

![](https://assets.underline.io/uploads/markdown_image/1/image/477a15e1897832130a1c615bcab204cb.jpg)

 Effects of Composition on the Ordered Phase Formation in Mn-Al and Mn-Ge Alloy Thin Films Grown on Cr(001) Single-Crystal Underlayers

The L10 structured τ-phase of the Mn-Al-C system gained interest for the potential use as a rare earth free hard magnetic material [1]. It has been reported that the formation of a small fraction of Mn3AlC precipitates leads to an increase in coercivity of the τ-phase [2]. Mn3AlC is also of interest because it is ferromagnetic, has a wide phase field in the Mn-Al-C equilibrium diagram and shares a two-phase region with the ε-phase of MnAl, which is the origin of τ-MnAl-C. Despite its broad homogeneity range, the Curie temperature (TC) of Mn3AlC is often reported to be at around 290 K [3,4]. There are only few reports of the influence of composition on the TC [5,6,7] and one has demonstrated a dramatic increase in TC up to 880 K with increasing Mn-content [5]. Furthermore, it has been reported that a full substitution of Al with Mn is possible, while retaining the ferromagnetic properties, resulting in Mn4C with a TC of 870 K [8]. The aim of the current work is to carry out a systematic investigation of the intrinsic magnetic properties of Mn3AlC in order to confirm the dramatic results presented in the literature and to assess the possible influence of Mn3AlC precipitates on the properties of MnAl-C permanent magnets. Two series of Mn-Al-C alloys have been prepared: in one series the Mn-content was constant while Al was substituted with C (compositions: Mn58Al34-xC8+x x = 0, 5, 10) and in the other series the C-content was constant, while Al was substituted with Mn (compositions: Mn60+xAl21-xC19 x = 0,5 ,10). The alloys were arc-melted in 10 g ingots, followed by a homogenisation at 1100°C in Ar-atmosphere for 5 days and were finally quenched into water. The chemical composition of the homogenised materials was determined using Inductively Coupled Plasma Optical Emission Spectroscopy for Mn- and Al-contents and by carrier-gas hot-extraction for C-content. In Fig. 1 and Fig. 2 the measured values of Mn-content were used for alloys produced in this study. Structural information about the phases in the alloys was gathered from XRD measurements, using a Bruker diffractometer with a Co-Kα radiation. Analysis of the XRD patterns was done using FullProf Suite software. Local chemical analysis of the microstructure was done by EDX-measurements using a Gemini Leo 1530 SEM. The magnetic properties of the alloys were measured using the VSM-mode of a Quantum Design PPMS. M(T)-curves were measured at an applied field of 0.1 T in a temperature range from 10 K to 900 K with a heating and cooling rate of 5Kmin-1. Furthermore, hysteresis loops up to an applied field of 14 T were measured at room temperature and at 100 K. In Fig.1 the TC of Mn3AlC is plotted against the Mn content of the phase from various sources. The Mn-content is related to the Mn3AlC formula unit containing 5 atoms. The curves represent the work of different groups that investigated the influence of composition on the TC. The results of the current work are shown in separate branches, since the two different series of alloys focus on different changes of the composition. The substitution of Al with C in the current work (solid black squares) shows a nearly constant TC at around 290 K, despite the change of the overall Mn-content. The same kind of substitution was done by Tong et al. [7], who reported a somewhat larger influence on TC: a change of 60 K was found in going from a composition of 2.75 to 3.0 Mn atoms per formula unit. Contrary to that, the substitution of Al with Mn in the current work (unfilled black squares) shows a drastic increase of the TC from around 355 K to 765 K. This behaviour of the TC is similar to the increase of the TC shown by the work of Wang et al. [5] and Butters et al. [6], who substituted Al with Mn as well. The work of Si et al. [8] focused on the properties of Mn4C, which shows a similar crystal structure to Mn3AlC. The TC of the Mn4C phase seems to fit well to the dependence of TC of Mn3AlC, which could confirm that a full substitution of Al with Mn is possible. Fig. 2 shows the magnetisation (M) as a function of Mn-content measured in the current work and data taken from Wang et al. [5]. The data for M was taken at 1.5 T, for which the samples are close to saturation. The Mn-content is related to the Mn3AlC formula unit. The M values are taken at 100 K. The curves represent the work of our group and the work of Wang et al. Compared to Fig. 1 we do not consider alloys of the first series, since they do not contain Mn3AlC only which means a change of M is connected to the phase fraction of Mn3AlC and not only connected to a change of composition. The alloy series in which Al was replaced by Mn shows a maximum of 92 Am2kg-1 at 3.4 Mn atoms per unit formula. Compared to the work of Wang et al. [5], who found a maximum for M of 90 Am2kg-1 for a stoichiometric composition the maximum of magnetisation is shifted to higher values of Mn atoms per unit cell. In the current work two different alloy series were used to study the effect of composition on the TC and M of Mn3AlC. The results confirm the dramatic increase in TC on substituting Al with Mn reported in the literature [5] and the decrease in M of Mn3AlC on increasing Mn content. The ferromagnetic nature of the Mn3AlC phase may influence the magnetic properties of τ-MnAl-C magnets, particularly for low temperature applications. To investigate this in more detail, further studies are needed.

**References:**

[1] Feng, L. et al. Acta Materialia 199, 155-168 (2020) [2] Ohtani, T. et al. IEEE Transactions on Magnetics 13, 1328–1330 (1977) [3] Dierkes, H. et al. Inorganic Chemistry 56, 1045-1048 (2017) [4] Wang, B. et al. Journal of Applied Physics 108, 093925 (2010) [5] Wang, X. et al. Journal of Magnetics 24 123-127(2019) [6] Butters, R. P.; Myers H.P. CI. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 46, 895-902 (1955) [7] Tong, P. et al. Solid State Communications 138, 64-67 (2006) [8] Si, P. et al. Applied Physcis Letter 112 192407 (2018)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/87556979e163a2e8decd39c3937fb997.jpg)
Fig. 1: Curie temperature of Mn3AlC depending on the Mn content per formula unit.

![](https://assets.underline.io/uploads/markdown_image/1/image/91b633b019279b6a999f1d50417249b0.jpg)
Fig. 2: Measured magnetisation at an applied field of 1.5 T at 100 K for different MnAlC-alloys.

 The Dependence of the Intrinsic Magnetic Properties of Mn3AlC on its Chemical Composition

L10-ordered MnAl shows a high magnetic moment of 2.4 μB/f.u. and magnetocrystalline anisotropy constant (K) of 1.5 MJ/m3 [1,2]. The magnetocrystalline anisotropy and spin direction can be tuned by either the axial ratio c/a or substituting the Mn of MnAl by transition metals. Sakuma calculated Ku of L10-ordered equiatomic MnAl as a function of the axial ratio c/a ranging from 0.60 to 1.2 by first-principles calculation [1]. It was found that the spin direction changes to the out-of-plane direction (uniaxial) from the in-plane direction at c/a = 0.7071 by changing the magnetocrystalline anisotropy sign to positive from negative. The calculated magnetocrystalline anisotropy constant is 1.5 MJ/m3 at c/a = 0.9058. Sakuma used the lattice constants of a =3.93 Å and c =3.56 Å measured at 300 K in his first-principles calculations. On the other hand, Manchanda et al. have performed the first-principles calculation on Fe-substituted L10-ordered MnAl to investigate magnetocrystalline anisotropy constant and found that Fe increases the magnetocrystalline anisotropy constant to 2.5 MJ/m3 from 1.5 MJ/m3 of non-substituted MnAl. However, the total magnetic moment decreased, and the spin direction did not change [3]. Neither Sakuma nor Manchanda reported the Curie temperature for various c/a MnAl and Fe-substituted MnAl and relaxed the studied MnAl system. In this paper, we have systematically investigated the effects of transition metals (Fe, Co, Ni) on magnetization, magnetocrystalline anisotropy, and Curie temperature of L10-ordered MnAl. We used the WIEN2k package based on density functional theory (DFT) and the full-potential linearized augmented plane wave (FPLAPW) method to conduct the first-principles calculations [4]. All calculations used 19 × 19 × 27 k-point mesh generating 1400 k-points in the irreducible part of the Brillouin zone to obtain electronic structures of Mn0.5TM0.5Al (TM = Mn, Fe, Co, and Ni). Each TM has the following valence electrons (n): 7 for Mn (3d5 4s2), 8 for Fe (3d6 4s2), 9 for Co (3d7 4s2), and 10 for Ni (3d8 4s2). After having relaxed L10-ordered Mn0.5TM0.5Al (TM = Mn, Fe, Co, and Ni), we obtained the lattice constants, volumes, and c/a ratios in Table 1. It is found that the c/a ratio and volume decrease with an increase in n from 7 (Mn) to 10 (Ni) in Table 1. Figure 1 shows the density of states (DOS) for Mn0.5TM0.5Al (TM = Mn, Fe, Co, and Ni) with spin configuration. DOS significantly changes with the TM substitution. It is striking that Ni-substituted MnAl shows the ferrimagnetic spin configuration, as seen in Fig. 1(d). However, the ferromagnetic spin configuration is set in Mn0.5TM0.5Al (TM = Mn, Fe, and Co). As summarized in Table 1, the total magnetic moment decreases as the valence electron increases in the order of Ni, Co, and Fe. In order to calculate the K of the Mn0.5TM0.5Al system, the magnetocrystalline anisotropy energy (MAE) was calculated using the total energy difference between &lt;001&gt; and &lt;100&gt; spin configurations (ΔE = E&lt;001&gt; - E&lt;100&gt;). The Fe substitution leads to a significant increase in K from 1.34 MJ/m3 to 2.98 MJ/m3. When either Co or Ni replaces Fe, the K becomes negative and is -0.3 MJ/m3 for Co or -0.18 MJ/m3 for Ni in Table 1. This implies the magnetization (Co, Ni) direction changes to the in-plane direction from the out-of-plane direction, i.e., uniaxial (Mn and Fe). Curie temperature (TC) was calculated by the mean-field approximation (MFA). The TC dramatically decreases from 685 K(Mn) to 20 K(Ni). Alloying the MnBi with transition elements could tune the spin direction and magnetocrystalline anisotropy of L10-ordered MnAl. However, magnetization and Curie temperature are negatively affected by the alloying. We will discuss all these findings in detail in this presentation. * This work was supported in part by the E. A. “Larry” Drummond Endowment at the University of Alabama.

**References:**

[1] A. Sakuma, J. Phys. Soc. Jpn. 63, 1422 (1994) [2] J. H. Park, Y. K. Hong, et al., J. Appl. Phys. 107, 09A731 (2010) [3] P. Manchanda, et al., J. Magn. Magn. Mater, 365, 88-92 (2014) [4] Blaha, P. and Schwarz, K., WIEN2k: An Augmented Plane Wave + Local Orbitals Program for Calculating Crystal Properties (Karlheinz Schwarz Techn. Universität Wien, Vienna, Austria, 2001)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/aecee2e75cceaee3a872a555081225de.jpg)
Table 1. Lattice parameters and magnetic properties of L10-Mn0.5TM0.5Al (TM = Mn, Fe, Co, Ni).

![](https://assets.underline.io/uploads/markdown_image/1/image/8c515420a3e4c27b6c54c6fa18f1b9d9.jpg)
Figure 1. Density of states (DOS) for (a) MnAl, (b) Mn0.5Fe0.5Al, (c) Mn0.5Co0.5Al, and (d) Mn0.5Ni0.5Al based relaxed lattice constants.

 Tuning Spin and Magnetocrystalline Anisotropy of L10-ordered MnAl with Transition Metal Elements (TM = Mn, Fe, Co, Ni)*

Nowadays the use of permanent magnets (PMs) plays a key role in different applications such as energy and electromobility. In order to obtain a high magnetic performance, the most extended choice is PMs containing rare-earths (REs). Alternative RE-free PM materials are investigated to plug the gap between ferrites and NdFeB magnets [1]. MnAlC alloy is a promising RE-free PM candidate, but further optimization is required in order to achieve a proper combination of saturation magnetization (provided by the only ferromagnetic phase of the system: L10 or τ-MnAl) and coercivity. Studies on MnAl alloys typically focus on obtaining a maximum content of τ-phase, which results in an optimized saturation magnetization [2]. We have recently shown the possibility of developing coercivity in gas-atomized MnAl powder by applying milling times as short as 30-270 s via the self-developed “flash milling” method [3]. A successful combination of nanostructuring and a controlled phase transformation process is behind the observed increase in coercivity [3]. In this study we have used ε-phase MnAlC powder produced by gas-atomization and flash-milling (60 s) to fabricate MnAlC bulk magnets, with no need to transform previously the ε into τ phase. Under the optimal combination of pressure (300 MPa) and temperature (600 °C), hot-pressing applied to ε-phase MnAlC alloy has been able to manage simultaneously the ε-to-τ transformation and the powder compaction to end with a high-density (93%) MnAlC magnet [4]. Figure 1(a) shows the X-ray diffraction pattern of the precursor ε-phase loose powder and the successful transformation into τ-phase after the hot-pressing process. The room temperature hysteresis loops measured for the hot-pressed magnet showed an enhancement of the coercivity of up to 25 % and a slightly superior magnetization at remanence by comparison with the annealed gas-atomized loose powder (Fig. 1 (b)). The large coercivity obtained for the magnet may be explained based on the formation of both β-Mn and Mn3AlC phases together with a reduced mean crystallite size (27 nm) and the induced strain during compaction [4]. On this basis, we have demonstrated that coercivity of the compacted magnet can be further increased in about 15% by starting from flash-milled gas-atomized powder, due to a larger content of β-Mn and Mn3AlC phases. We have obtained comparable magnetization and coercivity values when compacting the starting material with a crystallographic structure based on τ-phase (i.e., involving two steps: annealing for managing ε into τ phase transformation followed by hot-pressing), thus validating our single-step process. This is a promising cost-efficient route which opens the path to new possibilities for the fabrication of MnAlC-based magnets. Acknowledgements IMDEA acknowledges financial support from EU M-ERA.NET and MINECO through the projects "NEXMAG" (Project Success Case at H2020 M-era.Net Programme, Ref. PCIN-2015-126) and "3D-MAGNETOH" (Ref. MAT2017-89960-R); Regional Government of Madrid through "NANOMAGCOST" project (Ref. P2018/NMT-4321); and Höganäs AB through the industrial collaboration "ECNanoManga".

**References:**

[1] J.M.D. Coey, IEEE Trans. Magn., 47, 4671 (2011) [2] V. Øygarden et al., J. Alloys Compd., 779, 776 (2019) [3] J. Rial et al., Acta Mater., 157, 42 (2018); Engineering, 6 (2), 173 (2020) [4] C. Muñoz-Rodríguez et al., J. Alloys Compd., 847, 156361 (2020)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/ee90b2b1fd40d0e1617728e809b5f480.jpg)
Fig. 1. (a) X-ray diffraction pattern of the ε-phase gas-atomized loose powder and hot-pressed powder. (b) Room temperature hysteresis loops (second quadrant) of the loose annealed powder (ε transformed into τ-phase) and the magnet obtained by hot-pressing the ε-phase powder. Inset shows an image of the resulting compacted MnAlC magnet.

 MnAlC Permanent Magnets Obtained Directly from ε-Phase Gas-Atomized and Milled Powder by Hot-Pressing

Nowadays multi-material additive manufacturing (AM) is attracting much interest in many high technological sectors such as automotive, energy, electronics, aeronautics and medicine. The combination of AM technology with the synthesis of composite materials allows for designing and fabricating objects with complex shapes and tuned properties for reaching high performance [1]. For permanent magnet (PM) applications, a present challenge is developing magnets by AM with no geometrical constrictions, high filling factor (FF) to avoid diluting their magnetic properties, and no deterioration of their PM properties during processing [2]. The industrial fabrication of PM/polymer composites (typically in the shape of pellets) is done by mechanical methods. We showed recently, and for the very first time, the possibility of producing a continuous PM-based filament (meters long) using as a precursor a highly loaded PM/polymer composite synthesized by solution casting [3]. This technique allows for the synthesis of customized composites making possible the choice of the polymer according to the requirements dictated by the final application. Moreover, it is an easily scalable route, which opens a new path to the AM technology of PMs. In this study, different composites (PM particles/polymer) have been analyzed consisting on several PM materials: gas-atomized τ-MnAlC (Fig. 1(a)), Sr-ferrite, NdFeB, and hybrid (NdFeB/Sr-ferrite) powders embedded in different polymer matrix. It is of large scientific and technological interest to consider rare earth (RE)-free PM alternatives such as improved ferrites, and the promising MnAl-based alloys [4]. The starting powders showed different mean particle size ranging from the 5 μm of Sr-ferrite to the 50 μm of NdFeB powders. MnAlC powders showed an intermediate mean particle size between 15 and 30 μm. This work will show the effect of particle size (with special attention to the benefits of combining dissimilar sizes), polymer and fabrication parameters on the properties of the final products, showing that they are key factors to be considered and optimized for obtaining flexible and continuous filaments with a high FF. Particle size plays a key role when extruding the composites in order to obtain flexible filaments with a high FF [3,5,6]. It has been observed that if the mean particle size is larger than 20 μm (coarse particles), the extrusion process is affected by the composite rheology under the exerted pressures and, consequently, the FF of the extruded filament is reduced [5,6]. A key result from this study is the demonstration that, by mixing particles with different size and optimizing the fine-to-coarse particles ratio, it is possible to obtain flexible filaments with increased FF (up to 90%), i.e., leading to an enhanced effectiveness of the extrusion process [5,6]. MnAlC-based composites were synthesized by solution casting (Fig. 1(a)) making possible to tune the FF, reaching extremely high values above 85% [3]. Composites were extruded into continuous and flexible filaments with PM properties (Fig. 1(b)), with a length over 10 m [3,6]. The homogeneity of the composites and filaments has been determined by scanning electron microscopy, SEM (Fig. 1(b)). Processing of SEM images allowed for obtaining the FF of the composites and filaments. Vibrating sample magnetometry (VSM) has been used for accurately determining the FF of composites and filaments, being positioned as a faster technique in comparison with image processing. Moreover, and important in view of practical applications, this technique has demonstrated the no deterioration of PM properties of the starting particles after composite synthesis and filament extrusion processes (Fig. 1(b)) [3,5,6]. We will compare these results with those obtained in the preparation of filament prepared from composites based on: Sr-ferrite (Hc~3 kOe, FF=92%), NdFeB (Hc=10.2 kOe, FF=83%) and hybrid composites containing fine (Sr-ferrite) and coarse (NdFeB) particles (Hc=8 kOe, FF=90%). Optimized MnAlC-based filament (with a high MnAlC content above 80 wt.%) was used for 3D-printing objects as a proof-of-concept (Fig. 2). Magnetic measurements performed on the printed objects (Fig. 2) have proved that alternative PM materials can be efficiently synthesized and processed for developing novel PMs by AM under controlled processing temperature, which might be used in sensing devices [6]. Acknowledgements Authors acknowledge fruitful collaboration and discussions with B. Skårman, H. Vidarsson and P.-O. Larsson from Höganäs AB (Sweden), and A. Nieto and R. Altimira from IMA S.L.U. (Spain). Authors additionally acknowledge financial support from EU M-ERA.NET and MINECO through the projects "NEXMAG" (M-ERA.NET Project Success Case, Ref. PCIN-2015-126) and "3D-MAGNETOH" (Ref. MAT2017-89960-R); Regional Government of Madrid through "NANOMAGCOST" project (Ref. P2018/NMT-4321); and Höganäs AB through the industrial contract "GAMMA".

**References:**

[1] S.A.M. Tofail et al., Mater. Today, 21, 22 (2018) [2] C. Huber et al., Appl. Phys. Lett., 109, 162401 (2016) [3] E.M. Palmero et al., Sci. Technol. Adv. Mater., 19, 465 (2018) [4] A. Bollero et al., ACS Sustainable Chem. Eng., 5, 3243 (2017); J. Rial et al., Acta Mater., 157, 42 (2018); J. Rial et al. Engineering, 6 (2), 173; C. Muñoz-Rodríguez et al., J. Alloys Compd., 847, 156361 (2020) [5] E.M. Palmero et al., IEEE Trans. Magn., 55, 2101004 (2019) [6] E.M. Palmero et al., Addit. Manuf., 33, 101179 (2020)

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/07242cdfa3f27df187acf728ebcc1b22.jpg)
Fig. 1. (a) Composite synthesis by solution casting with representative images of the starting MnAlC powders and the resulting MnAlC/polymer composite. (b) Second quadrant of the hysteresis loops measured at room temperature for the gas-atomized MnAlC particles, MnAlC/polymer composite and extruded filament. Inset shows an extruded MnAlC-based filament together with an SEM image of its circular cross section.

![](https://assets.underline.io/uploads/markdown_image/1/image/313f4a31e639429f97217f5044912754.jpg)
Fig. 2. Magnetic pieces obtained by additive manufacturing using an extruded filament based on MnAlC particles, together with the magnetic flux density measured at the surface of a 3D-printed disc.

 Additive Manufacturing of Rare Earth-free and Hybrid Permanent Magnets: from Composites synthesized by Solution Casting to Magnetic Filament and 3D-printing of Magnets

In recent years, magnetic nanocomposites (NCs) have obtained an ever-increasing attention, due to the possibility to finely control and modify their features at the nanoscale, which is fundamental for extending their applicability as permanent magnets in a multitude of energy-related technological areas [1,2]. In this regard, ferrites have shown to be a promising class of materials, in particular M-type SrFe12O19 (SFO) and spinel CoFe2O4 (CFO) nanoparticles (NPs): their respective hard and soft magnetic properties make them interesting candidates for hard-soft exchange-coupled nanostructures. Effectively, the combination of two prototypical phases with different extrinsic properties expands the potentiality and hence the suitability of such nanosystems [3]. Our main focus here is on the investigation of the synthesis strategy in controlling the magnetic coupling, through the optimization of the interface between the hard-soft magnetic phases, thus achieving efficiently strongly coupled nanostructures. This work addresses a systematic investigation of the magnetic interaction in SFO-CFO hard-soft NCs. Several samples were synthesized through a self-combustion sol-gel approach, with compositions ranging from 50/50 to 90/10 w/w %, in step of 10%. In order to elucidate the synthesis method, a thermogravimetric/differential thermal analysis (TG/DTA) was performed, showing a first evidence that the reaction proceeds in a symbiotic way, where the formation of CFO after-combustion assists the evolution of SFO precursors, when the CFO % is increased [4]. The subsequent investigation on the inner structure of the NCs, by transmission electron microscopy (TEM) technique, revealed that SFO NPs are in the form of platelets, with CFO nanocrystals confined between the SFO ones. This clearly suggests a strong interaction between the two structures [5]. The evidence of a symbiotic growth was furtherly confirmed by X-ray powder diffraction (XRPD): the crystallite size extracted by Rietveld refinement shows an evident dependence of SFO crystallite size on the amount of inserted soft phase. When CFO increases, the dc (size extracted along the c-axis) and dab (size extracted along the ab-plane) for SFO drop from 104 to 51 nm from 132 to 77 nm, respectively. Hence, we illustrate here for the first time that the CFO grains act as nucleation sites for the SFO phase, resulting in an effect of confinement, which restricts the growth of one phase by the other. To clarify the relationship between the morphological-structural features and magnetic coupling, the static magnetic properties of NCs were investigated at 300 K using a superconducting quantum interference device (SQUID) magnetometer. The analysis of field-dependent magnetization loops shows that the two magnetic phases are homogeneously dispersed and strongly coupled (as reported in Figure 1). Furthermore, the switching field distributions clearly exhibit a single reversal process of magnetization, confirming that the SFO and CFO phases are strongly coupled in the composite (as shown in Figure 2). As expected, the variation of coercivity (HC) is consistent with the increasing amount of soft CFO phase, as it decreases from ~463 to 178 kA/m, within the limit of exchange-coupling, whereas the saturation moment (MS) increases up to ~74 Am2/kg. Our study shows that it is possible to control and limit the decrease of the coercivity of hard-soft bi-magnetic NCs to our desire by controlling the size and distribution of the hard-soft regions, and their interface. We thank the Swedish Energy Agency and Swedish Research Council (VR) for financially supporting this work.

**References:**

[1] R. Skomski, P. Manchanda, P. K. Kumar, B. Balamurugan, A. Kashyap, D. J. Sellmyer, IEEE Trans. Magn. , 49 (7), 3215, (2013) [2] K. P. Skokov, O. Gutfleisch, Scr. Mater., 154, 289, (2018) [3] G. Muscas, P. Anil Kumar, G. Barucca, G. Concas, G. Varvaro, R. Mathieu, D. Peddis, Nanoscale, 8, 2081, (2016) [4] F. Sayed, G. Kotnana, G. Muscas, F. Locardi, A. Comite, G. Varvaro, D. Peddis, G. Barucca, R. Mathieu, T. Sarkar, Nanoscale Adv., 2 (2), 851, (2020) [5] P. Maltoni, T. Sarkar, G. Varvaro, G. Barucca, S. Ivanov, D. Peddis, R. Mathieu, J. Phys. D: Appl. Phys., in print, (2020), doi:10.1088/1361-6463/abd20d,

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/216005caea78a23890b0a81308d7264d.jpg)
Fig.1 Magnetization M vs. magnetic field H curves for SFO, SFO-CFO 50/50 w/w % and CFO at 300K

![](https://assets.underline.io/uploads/markdown_image/1/image/9b44e3133f0dd8474dbde85d43ae2b6b.jpg)
Fig.2 Normalized irreversible susceptibility χirr vs reverse magnetic field H for SFO, SFO-CFO 50/50 w/w % and CFO at 300K; the inset shows the MDCD vs H.

 Morphology, structure and magnetic coupling relationship in hard-soft SrFe12O19-CoFe2O4 nanostructures

Permanent magnets are receiving an increasing economical impact due to their role in the Low Carbon Green Transition being employed for energy generation (wind generators), energy storage (flywheels) and electric transport. However, the most performing magnets, composed by rare earth (RE) elements, are in the head of the list of the Critical Raw Materials due to the recent uncertainty in supply and the increase of the costs of the RE. This last question has pushed the research of alternatives to RE magnets. Between the different solutions, the improvement of the performances of the hard ferrites is receiving an increasing interest [1-4]. In fact, ferrite magnets constitute by far the most widely used permanent magnet by weight [2,3] and any improvement of the performances is expected to have an important impact [4]. The most promising approach to improve the performances of ferrite magnets is the development of nanocomposites formed by magnetically coupled high magnetization and hard magnetic nanophases [4]. We present a research on a novel class of hybrid magnet composed by micrometric nanowires (NWs) and micrometric ferrites that exhibit improved hard magnetic properties. We have investigated the properties of composites containing micrometric FeCo nanowires and micrometric SrFe12O19 powders [5] to produce bonded magnets. FeCo NWs were prepared by template-assisted electrochemical deposition using nanoporous templates and polycarbonate nanoporous membranes to obtain NWs with different diameters (30, 50, and 100 nm). FeCo NWs−strontium ferrite mixture composites in the powder form were fabricated by mixing each phase. The evolution of the magnetic properties of the mixtures considering the different diameter of the NWs and the soft-hard ratio will be shown. The coercive field of the composites decreases as the soft contain is larger, while remanence magnetization remains almost constant but also increases depending on NWs size. The magnetic coupling between the two phases has been investigated using remanence measurements. Even the soft nature of the NWs, the hard and soft phases are magnetically coupled up to a 20%wt of the soft phase in the case of the NWs of 50 nm of diameter. A 20% increase in remanence together with a weak decrease in coercivity, as consequence of the magnetodipolar interaction between ferrite particles and FeCo NWs, is observed. In fact, the oriented bonded composites magnets exhibit a larger energy product up to 40% larger than the corresponding ferrite ones. The magnetic properties of these magnets will be discussed considering the magnetic configuration of the Nanowires. This research was supported by EU- H2020 AMPHIBIAN Project (H2020-NMBP-2016-720853) and by the JECS Trust.

**References:**

[1] EU-H2020, AMPHIBIAN Project. Website : amphibianproject.eu. [2] R. C. Pullar, Progress in Materials Science 57, 1191 (2012). [3] S H Mahmood and I Abu-Aljarayesh Hexaferrite Permanent Magnetic Materials. Materials Research Forum LLC, (2016) [4] C. de Julián Fernández et al., J. Applied Phys. D: Applied Physics, (2020) in press

**Images:**

 Metal Nanowire−Strontium Ferrite Composites for Free Rare-Earth Magnets

The design and development of alloys are typically based on one or two main constituents since the Bronze age, whereby the search for new materials using this millennia-old technique is reaching limits. Today, a new class of materials, using a different design approach of alloying multiple elements (five or more) to form compositions without dominant elements and therefore having high entropy of mixing, can discover materials with excellent properties, some of them being superior or unique compared to the conventional materials. This is also known as the high-entropy alloy (HEA) design concept where its multi-principal-element composition encompasses a vast compositional space, which further indicates many opportunities for materials design. While HEA research expands at a rapid growth since its first report in 2004 and has evolved from 1st to 2nd generation HEAs, most of the efforts focus on structural applications while their functional reports are scarce in comparison to their mechanical properties. Furthermore, when compared to conventional functional materials, they underperform due to their modest functional properties. One of their most reported functional properties, the magnetocaloric effect (MCE), only shows large values among HEAs for those compositions containing rare-earth elements. The rare-earth-free HEAs, on the other hand, typically exhibit very small MCE instead. In this talk, we will show that it is possible to enhance the MCE of rare-earth-free HEAs by at least one order of magnitude upon introducing a magneto-structural phase transition: isothermal entropy change increases from 1.7 to 13.1 J kg-1 K-1 (for 2.5 T) [1,2]. The significant improvement in the MCE performance is further shown in Fig. 1 when comparing the isothermal entropy change and the temperature averaged entropy change, TEC(10), to the literature. Instead of sampling the huge HEA compositional space by brute force, we have followed a directed search procedure to develop FeMnNiGexSi1-x HEAs that enabled us to put the functionality of HEAs comparable to some of the high-performance traditional materials. The variation of Ge/Si ratio enables the tuning of the thermomagnetic phase transitions (magneto-structural transformation and Curie transitions), leading to a further enhancement in the MCE. Our work is also found to merge the gap between magnetocaloric HEAs versus conventional magnetocaloric materials. These findings demonstrate the promising potential of HEAs for magnetocaloric applications. Work supported by AEI/FEDER-UE (grant PID2019-105720RB-I00), US/JUNTA/FEDER-UE (grant US- 1260179), Consejería de Economía, Conocimiento, Empresas y Universidad de la Junta de Andalucía (grant P18-RT-746), Army Research Laboratory under Cooperative Agreement Number W911NF-19-2-0212.

**References:**

[1] J.Y. Law, L.M. Moreno-Ramírez, Á. Díaz-García, A. Martín-Cid, S. Kobayashi, S. Kawaguchi, T. Nakamura, V. Franco, J. Alloy. Compd. 855 pp. 157424 (2021) . [2] J.Y. Law, Á. Díaz-García, L.M. Moreno-Ramírez, V. Franco, Submitted (2021). [3] M.C. Gao, D.B. Miracle, D. Maurice, X. Yan, Y. Zhang, J.A. Hawk, J. Mater. Res. 33(19) pp.3138-3155 (2018).

**Images:**

![](https://assets.underline.io/uploads/markdown_image/1/image/c0853824f7a3d0b587ae32052421cf30.png)
Fig 1. Performance comparison of magnetocaloric HEAs for 2.5 T based on isothermal entropy change peak, its corresponding temperature and temperature averaged entropy change over a 10 K span, TEC(10). Each data point represents a single material, with a diameter size proportional to its TEC(10). Literature data are collected from refs. [1 - 3].

Downloads

Next from INTERMAG 2021

Fcc-Co Clusters in L10-FePt Matrix as Graded-Interface Magnetic Nanocomposite

Similar lecture

System Level Sequential Subspace Design Optimization of Switched Reluctance Motor Drive Systems Based on Space Reduction Strategy

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from INTERMAG 2021

Fcc-Co Clusters in L10-FePt Matrix as Graded-Interface Magnetic Nanocomposite

Similar lecture

System Level Sequential Subspace Design Optimization of Switched Reluctance Motor Drive Systems Based on Space Reduction Strategy

Downloads