Development of Heusler-alloy. based magnetocaloric inks for 2D - 3D printing

DOCTORAL THESIS DEVELOPMENT OF HEUSLER – ALLOY – BASED MAGNETOCALORIC INKS FOR 2D – 3D PRINTING BOSCO RODRIGUEZ CRESPO 2024 Supervisors: Dr. Daniel Salazar Prof. Volodymyr Chernenko (cc)2024 BOSCO RODRIGUEZ CRESPO (cc by 4.0) Este trabajo de investigación para optar al Grado de Doctor por la Universidad del País Vasco (UPV/EHU) se ha realizado en colaboración entre el Basque Center for Materials, Applications and Nanostructures (BCMaterials), el Departamento de Electricidad y Electrónica de la Facultad de Ciencia y Tecnología y el instituto Funktionale Materialien (FM) de TU Darmstadt. Por otro lado, la tesis ha sido posible gracias a la financiación obtenida desde diferentes fuentes a nivel nacional y estatal, como la beca para la formación de personal investigador con instituciones y empresas (PIFI20/10) otorgada por la UPV/EHU. Asimismo, esta tesis también ha sido posible gracias a la ayuda financiera del Ministerio de Ciencia, Innovación y Universidades (Proyecto RTI2018-094683B-C53-54) y el Departamento de Educación del Gobierno Vasco (Proyecto IT1245-19). También agradecer a la financiación del Departamento de Industria y Educación del Gobierno Vasco mediante los programas ELKARTEK y PIBA (PIBA-2018-06). I Agradecimientos Cuando empecé este viaje hace poco más de cuatro años, nunca imaginé la profundidad del compromiso y las adversidades que encontraría en el camino. La realización de esta tesis doctoral ha sido una travesía llena de desafíos y descubrimientos, y hoy, al llegar al final de este capítulo, me siento profundamente agradecido por todas las personas que han sido parte de esta experiencia. Lo más difícil de escribir estas líneas va a ser no olvidar de mencionar a todas las personas, pues en esta etapa son muchas con las que he tenido la fortuna de compartir experiencias y conocimientos. En primer lugar, quiero expresar mi agradecimiento a mis directores de tesis Dr. Daniel Salazar y Prof. Volodymyr Chernenko, pues sin ellos ni su apoyo nada de esto hubiera sido posible. Su guía experta, apoyo inquebrantable y confianza en mí han sido fundamentales para este logro. Han sido no solo mentores, sino también modelos a seguir que me han inspirado a alcanzar mis metas académicas. Agradecer también a mi tutor Jon Gutiérrez por su ayuda. Quiero agradecer también a la Fundación BCMaterials por la gran oportunidad que me dieron para poder realizar allí la tesis y a la Universidad del País Vasco (UPV/EHU) por otorgarme la beca para la formación de personal investigador con instituciones y empresas (PIFI20/10). A mi familia, a quienes debo tanto, quiero agradecerles por su apoyo constante. A mis padres, por inculcarme valores y proporcionarme la base sólida desde la que he podido construir mi carrera profesional y personal. En el laboratorio hemos acabando formando un equipo excepcional de compañeros y, sobre todo amigos, entre Andrés G, James R, Mariana R, Santiago C y los numerosos visitantes y colaboradores que nos han acompañado a lo largo de estos años. Cada uno ha aportado su granito de arena a que cada II día ir a trabajar sea más ameno. Las discusiones intelectuales y no tan intelectuales en los cafés diarios (la razón real por la que todos venimos al trabajo) y las colaboraciones que hemos forjado han sido esenciales para hacer que la tesis sea mucho más disfrutable. La tesis ha traído consigo una sucesión de muchos altibajos los cuales han sido sobrellevados mucho mejor gracias a la compañía y motivación de mis compañeros. En especial, mi grupo de metaleros metalúrgicos, J Napal, A Reizábal, D Payno, P González, A García, F Zheng, N Barroso, P Lazpita, JM Porro, N Perinka, S Lanceros, el equipo de administración de BCMaterials y todos los compañeros con los que he coincidido y colaborado. I also want to thank to all the people of Funktionale Materialien (FM) from TU Darmstadt, that allowed me to make a three month stay in their center, where I had the opportunity to work with people that are a world reference for the subject of my thesis and that let me work in their facilities. In special, thank you Prof. Dr. O Gutfleisch, Dr. K Skokov, Dr. F Scheibel, Dr. A Aubert, Dr. S Ener, K Schäfer, B Beckmann, Rafael G, Dr. L Pfeuffer, Dr. L Schäfer, M Laux and all the people that made my stay unforgettable. También quiero agradecer al personal de servicios generales de la universidad, por su gran apoyo y ayuda a la hora de poder hacer uso de sus equipos. A Iñaki Orue, por las eternas peleas con el VSM y el SQUID y a Aitor Larrañaga por su gran ayuda a la hora de hacer medidas de difracción de rayos X. Y, por último, pero no menos importante, a Dídac Mesa Romeu, por ser el hermano mayor que nunca tuve, y por permitirme usar algunas de sus increíbles astrofotografías nocturnas para la contraportada de la tesis y algunos capítulos. Gracias por haber sido parte integral de este capítulo de mi vida. Con aprecio sincero, Bosco. III Resumen En la búsqueda de tecnologías de refrigeración sostenibles y eficientes en energía, el estudio de los efectos calóricos ha surgido como una vía prometedora. Esta tesis doctoral se embarca en una exploración extensa del efecto magnetocalórico, con el objetivo principal de permitir la impresión 3D rentable de estructuras magnetocalóricas utilizando materiales respetuosos con el medio ambiente. La investigación abarca la síntesis de aleaciones Heusler magnéticas con memoria de forma, la preparación de cintas metálicas, tratamientos térmicos, producción de polvo y la transformación de estos materiales en tintas imprimibles, culminando en la fabricación aditiva de estructuras 3D complejas con propiedades magnetocalóricas conservadas en cada paso. La tesis comienza con una introducción exhaustiva de los efectos calóricos, con un enfoque en el efecto magnetocalórico. El efecto magnetocalórico, caracterizado por cambios de temperatura en respuesta a variaciones en el campo magnético, representa una vía alternativa para soluciones de enfriamiento sostenible y gestión térmica eficiente en energía. Entre los materiales existentes que muestran el efecto magnetocalórico, la tesis se centra en una familia de materiales llamada aleaciones magnéticas de memoria de forma Heusler, específicamente en aquellas basadas en combinaciones de Níquel y Manganeso con otros elementos. El primer paso de la tesis es proporcionar una visión general completa de los materiales magnetocalóricos del tipo Heusler, con un enfoque en sus propiedades estructurales y magnéticas. Esta sección incluye una exploración detallada del efecto magnetocalórico, las transiciones de fase y los principios subyacentes de estos materiales. El estudio profundiza en la incorporación de ciertos elementos en los sistemas de aleaciones, elucidando su impacto IV profundo en las propiedades de las aleaciones Heusler y la mejora del efecto magnetocalórico. El trabajo se centra en lograr composiciones de aleación óptimas que equilibren la eficiencia y las consideraciones ambientales. La investigación pasa de la exploración teórica basada en la investigación de la literatura a la implementación práctica a medida que se fabrican múltiples aleaciones magnetocalóricas de Heusler mediante diferentes técnicas. Empleando la técnica de melt-spinning, las aleaciones se transforman en cintas metálicas, allanando el camino para la optimización posterior a través de tratamientos térmicos estratégicamente diseñados. Se realiza un estudio sistemático de los tratamientos térmicos para afinar las propiedades magnéticas y magnetocalóricas de las cintas, con el fin de lograr las propiedades óptimas. Un avance innovador en la tesis gira en torno a la transformación de las cintas en tintas imprimibles para su posterior implementación en impresión 2D y 3D. A través de molienda mecánica, las cintas se convierten en polvo, al cual se le realizan tratamientos térmicos para contrarrestar los efectos de la degradación del material inducida por la molienda. Este polvo más adelante se mezcla con diversos polímeros biodegradables, entre ellos celulosa, que es un polímero respetuoso con el medio ambiente y con mucha biodisponibilidad, para crear tintas imprimibles para aplicaciones de fabricación aditiva, inaugurando la capacidad de elaborar estructuras 3D, preservando las propiedades magnetocalóricas de las cintas. Para desarrollar tanto la técnica de impresión 2D, mediante serigrafía, como la de 3D, mediante el proceso llamado impresión por extrusión, primeramente, se emplearon polvos metálicos comerciales (Hierro, Aluminio, Silicio) sin ninguna funcionalidad, dado que estos polvos son económicos y fáciles de obtener. De esta forma se abrió la puerta a, primeramente, desarrollar la ruta de creación y optimización de tintas imprimibles y, después, a encontrar las limitaciones de V las técnicas de impresión, y encontrar las estructuras que eran imprimibles y encontrar las limitaciones que cada técnica de impresión tenía. Una vez encontrados los parámetros que optimizaban los procesos de impresión se implementó a los polvos magnetocalóricos para así obtener films 2D y estructuras 3D magnetocalóricas de diversas formas y tamaños. La primera validación de las tintas magnetocalóricas se realizó mediante impresión 2D por serigrafía. Las primeas pruebas de impresión se realizan imprimiendo figuras geométricas muy simples de una sola capa y posteriormente se aplican múltiples capas para aumentar así la cantidad de material funcional que contiene el film impreso. Debido a la naturaleza de la celulosa, se obtuvieron films flexibles con un alto contenido en material magnetocalórico, que podrían servir para aplicaciones en dispositivos de electrónica. Además, estos films impresos preservan las propiedades magnetocalóricas del material funcional precursor, quedando así validad la técnica de impresión 2D para materiales magnetocalóricos (ver Figura 1). Figure 1: Esquema del proceso de impresión 2D por serigrafía. Las propiedades magnetocalóricas del film flexible impreso son análogas a las de las cintas, que constituyen el material precursor. XII XIII Abstract In the quest for sustainable and energy-efficient cooling and refrigeration technologies, the study of caloric effects has emerged as a promising avenue. This doctoral thesis embarks on an extensive exploration of the magnetocaloric effect, with a primary goal of enabling the cost-effective 2D – 3D printing of magnetocaloric structures using environmentally-friendly materials. The research encompasses the synthesis of Heusler magnetic shape memory alloys, ribbon preparation, heat treatments, powder production, and the transformation of these materials into printable inks, culminating in the additive manufacturing of complex 2D – 3D structures with retained magnetocaloric properties at each step. An in-depth exploration into the magnetocaloric effect inherent in Heusler magnetic shape memory alloys, with a primary focus on elucidating the potential technological applications of this caloric phenomenon is studied. The introductory discourse provides a thorough examination of caloric effects, with specific attention directed towards the magnetocaloric effect and its relevance in diverse technological domains. The research rigorously investigates the off-stoichiometric NiMn-X (X=Sn, In,Ga) Heusler alloys, elucidating the nuanced impact of the incorporation of dopant elements such as Cobalt, Copper and Iron on their transformation, magnetic and magnetocaloric properties. Employing the melt-spinning technique, ribbons are fabricated and subjected to a systematic heat treatment protocols designed to optimize their transformation, magnetic and magnetocaloric properties. A route for powder preparation from the ribbons is studied and established. This powder undergoes a refined heat treatment procedure to mitigate the XIV detrimental effects of grinding, ensuring the preservation of magnetocaloric properties during this material processing. Characterization of both ribbons and powder serves as the foundational groundwork for the subsequent exploration of additive manufacturing applications. Formulation of printable inks utilizing cellulose as a polymer facilitates the creation of novel 2D-3D structures through additive manufacturing techniques. Following the printing process, the fabricated structures undergo a meticulous heat treatment regime, involving calcination to remove the polymer and sintering to achieve structural compactness and mechanical integrity. Various sintering routes are systematically investigated to discern the optimal approach that balances mechanical robustness with the preservation of magnetocaloric effects throughout each stage. The overarching objective of this research is to demonstrate the sustained retention of magnetocaloric properties across the entire spectrum of processes, from alloy fabrication to melt-spinning, heat treatment to grinding, additive manufacturing to sintering and getting 100% magnetocaloric 3D structure. The findings not only contribute nuanced insights into the manipulation of Heusler alloys for practical applications but also pave the way for the development of advanced materials with enhanced magnetocaloric functionalities in the sector of additive manufacturing. XV Visual Abstract XVI XVII Contents Agradecimientos I Resumen III Laburpena IX Abstract XIII Visual Abstract XV 1. Chapter 1: Introduction............................................................................1 1.1. Motivation.....................................................................................................3 1.2. Caloric Materials.........................................................................................4 1.2.1. Magnetocaloric effect..........................................................................5 1.2.1.1. Conventional magnetocaloric effect......................................8 1.2.1.2. Inverse magnetocaloric effect................................................13 1.2.1.3. Refrigeration capacity.............................................................13 1.2.2. Elastocaloric effect.............................................................................15 1.3. Martensitic transformation and shape memory effect..............16 1.4. Heusler-type NiMn-based magnetic shape memory alloys (MMMAs)...............................................................................................20 1.4.1. NiMn-based Heusler alloys...............................................................22 1.4.1.1. Effects of doping.....................................................................25 (a) Co addition....................................................................................25 (b) Cu addition....................................................................................27 1.4.1.2. Effect of fabrication methods...............................................28 1.4.1.3. Heat treatments.......................................................................29 1.5. Applied aspects of magnetocaloric materials..........................31 1.5.1. Active Magnetic regenerators............................................................31 1.5.2. Magnetic refrigerator..........................................................................32 1.6. Additive Manufacturing.................................................................34 1.6.1. Screen-printing....................................................................................35 1.6.2. Cold extrusion printing......................................................................36 1.7. State of the art....................................................................................37 1.8. Objectives ..........................................................................................38 1.9. References..........................................................................................39 XVIII 2. Chapter 2: Experimental Methods....................................................47 2.1. Choosing alloy composition.........................................................49 2.2. Alloys Fabrication............................................................................49 2.2.1. Induction casting................................................................................49 2.2.2. Arc-melting..........................................................................................50 2.3. Melt-Spinning...................................................................................51 2.4. Powder Preparation.........................................................................52 2.4.1. Manual grinding..................................................................................52 2.4.2. Hammer milling..................................................................................52 2.4.3. Gas atomization.................................................................................53 2.5. Heat Treatments..............................................................................53 2.6. Ink Preparation.................................................................................54 2.6.1. Collagen-based ink..............................................................................54 2.6.2. Silk-based ink.......................................................................................55 2.6.3. Cellulose-based ink.............................................................................56 2.7. 2D Printing ........................................................................................57 2.7.1. Doctor Blade.......................................................................................57 2.7.2. Screen-printing....................................................................................57 2.8. 3D Printing ........................................................................................58 2.9. Characterization Methods.............................................................59 2.9.1. Vibrating Sample Magnetometer (VSM).........................................59 2.9.2. Superconducting Quantum Interference Device (SQUID)..........60 2.9.3. Differential Scanning Calorimetry (DSC.........................................60 2.9.4. Scanning Electron Microscopy (SEM).............................................61 2.9.5. Magnetic field-induced adiabatic temperature change...................61 2.9.6. Mechanical characterization..............................................................62 2.9.7. X-Ray Diffraction...............................................................................62 2.9.8. Magnetic field induced entropy change calculation.......................63 2.10. References..........................................................................................64 3. Chapter 3: Search for NiMn-Based MSMAs in Ribbon Form with Potentially Enhanced MCE Performance..........................65 3.1. Introduction ......................................................................................67 3.2. NiMnSn MSMA System.................................................................69 3.2.1. Mn42.5Ni40Co8Sn9.5...............................................................................69 3.2.1.1. Composition and microstructure..........................................69 3.2.1.2. Transformation characteristics..............................................70 XIX 3.2.2. Mn48Ni35.5Sn8Co6.5Fe2 .........................................................................72 3.2.2.1. Composition and microstructure..........................................72 3.2.2.2. Transformation characteristics..............................................73 3.2.2.3. “Magnetic field – temperature” phase diagrams of martensitic transformation......................................................................74 3.2.2.4. Magnetic field induced entropy change................................75 3.2.2.5. Refrigerant capacity.................................................................76 3.2.2.6. Adiabatic magnetocaloric effect............................................77 3.2.3. Ni43Mn39Co7Sn11 .................................................................................79 3.2.3.1. Composition analysis..............................................................79 3.2.3.2. Transformation characteristics..............................................79 3.3. NiMnGa MSMA System................................................................81 3.3.1. Ni49Mn20Cu6Ga23Fe2...........................................................................81 3.3.1.1. Composition and microstructure..........................................81 3.3.1.2. Transformation characteristics..............................................82 3.3.2. Ni46Mn30Co5Ga20 ................................................................................83 3.3.2.1. Composition and microstructure..........................................83 3.3.2.2. Transformation characteristics..............................................83 3.3.3. Ni46Mn31Co5Ga17Fe1...........................................................................85 3.3.3.1. Composition and microstructure..........................................85 3.3.3.2. Transformation characteristics..............................................86 3.3.4. Ni50Mn18.7Cu6.25Ga25 ...........................................................................87 3.3.4.1. Composition and microstructure..........................................87 3.3.4.2. Transformation characteristics..............................................88 3.3.5. Ni50Mn18Cu5Ga25Fe2...........................................................................89 3.3.5.1. Composition analysis..............................................................89 3.3.5.2. Transformation characteristics..............................................89 3.4. NiMnIn MSMA System.................................................................90 3.4.1. Ni45.2Mn36.7Co5.1In13.0..........................................................................90 3.4.1.1. Composition analysis..............................................................90 3.4.1.2. Transformation characteristics..............................................91 3.5. Conclusions.......................................................................................92 3.6. References..........................................................................................93 4. Chapter 4: Production and Investigation of NiMn-Based MSMAs Powders with Enhanced MCE Properties..................95 4.1. Powder production from ribbons and by gas atomization...97 4.2. NiMnSn MSMA System.................................................................99 XX 4.2.1. Mn42.5Ni40Co8Sn9.5...............................................................................99 4.2.1.1. Composition and microstructure..........................................99 4.2.1.2. X-ray diffraction......................................................................99 4.2.1.3. Transformation characteristics............................................102 4.2.2. Mn48Ni35.5Sn8Co6.5Fe2.......................................................................103 4.2.2.1. Transformation characteristics............................................103 4.2.3. Ni43Mn39Co7Sn11 ...............................................................................104 4.2.3.1. Composition and microstructure........................................104 4.2.3.2. Transformation characteristics............................................105 4.2.3.3. “Magnetic field – temperature” phase diagrams of martensitic transformation...................................................................107 4.2.3.4. Magnetocaloric effect...........................................................110 4.2.3.4.1. Magnetic field-induced entropy change.................110 4.2.3.4.2. Refrigeration capacity...............................................112 4.2.4. Ni49.8Mn36.6Sn13.6 ..............................................................................113 4.2.4.1. Composition and microstructure........................................113 4.2.4.2. Transformation characteristics............................................114 4.2.4.3. “Magnetic field – temperature” phase diagrams of martensitic transformation...................................................................114 4.2.4.4. Magnetocaloric effect...........................................................115 4.2.4.4.1. Magnetic field-induced entropy change.................115 4.3. NiMnGa MSMA System..............................................................117 4.3.1. Ni50Mn18.7Cu6.25Ga25........................................................................117 4.3.1.1. Composition and microstructure........................................117 4.3.1.2. Transformation characteristics............................................117 4.4. Conclusions.....................................................................................118 4.5. References........................................................................................119 5. Chapter 5: Design and Fabrication of Novel Metallic Printable Materials...................................................................................121 5.1. 2D and 3D printing of commercial powders..........................123 5.1.1. Technique validation........................................................................123 5.1.2. Searching for a binder and solvents for eco-friendly approach using metallic powder as filler..............................................................124 5.1.2.1. Silk-based ink.........................................................................124 5.1.2.2. Collagen-based ink................................................................124 5.1.2.3. Cellulose-based ink...............................................................124 5.1.3. Commercial powders and ink parameters....................................125 XXI 5.1.4. Printing commercial powders.........................................................127 5.1.4.1. Screen-printing......................................................................127 5.1.4.2. Cold-extrusion printing.......................................................129 5.1.4.2.1. Iron (Fe) ink...............................................................130 5.1.4.2.2. Aluminium (Al) ink...................................................132 5.1.4.2.3. Silicon (Si) ink.............................................................133 5.1.5. Heat treatments: Calcination and sintering...................................134 5.1.6. Nickel electrodeposition..................................................................136 5.1.7. SEM characterization: Composition topography.........................137 5.1.8. Mechanical characterization............................................................141 5.2. 2D And 3D Printing of Magnetocaloric Heusler-Type MSMAs.........................................................................................................142 5.2.1. Silk-based MCE ink......................................................................142 5.2.1.1. Ink preparation and printing tests.......................................142 5.2.1.2. Transformation characteristics and magnetocaloric effect........................................................................................................143 5.2.2. Cellulose-based MCE inks..........................................................144 5.2.2.1. Ink parameters.......................................................................144 5.2.3. Screen-printing of Mn42.5Ni40Co8Sn9.5 magnetocaloric ink.145 5.2.3.1. Printing tests..........................................................................145 5.2.3.2. Magnetic anisotropy in printed samples.............................146 5.2.3.3. “Magnetic field – temperature” phase diagrams of martensitic transformation...................................................................147 5.2.3.4. Magnetocaloric effect...........................................................149 5.2.3.4.1. Magnetic field induced entropy change..................149 5.2.3.4.2. Adiabatic magnetocaloric effect..............................152 5.2.4. 3D cold-extrusion printing of magnetocaloric inks.............153 5.2.4.1. Mn42.5Ni40Co8Sn9.5...............................................................153 5.2.4.1.1. Printing.......................................................................153 5.2.4.1.2. Microstructure and composition analysis..............154 5.2.4.1.3. Calcination and sintering: Transformation characteristics....................................................................................155 5.2.4.1.4. Magnetocaloric effect................................................157 5.2.4.2. Ni49.8Mn36.6Sn13.6...................................................................159 5.2.4.2.1. Printing tests...............................................................159 5.2.4.2.2. Calcination and sintering..........................................160 5.2.4.2.3. Microstructure analysis.............................................160 5.2.4.2.4. Transformation characteristics................................161 6 Introduction later in 1878[9]. Thomson deduced that iron would warm up when a magnetic field was applied to it and cool down when the field was removed. While some literature claims that E. Warburg was first who has observed the MCE in iron in 1881[10][11], the phenomenon was explicitly demonstrated only in 1917 by Weiss and Piccard. They measured Ni around its Curie temperature at 1.5T and found a temperature change of 0.7 K[11]. In 1933, Giauque and MacDougall achieved sub Kelvin temperature of 0.25 K using a demagnetization of the paramagnetic Gd2(SO4)3 × 8H2O salt cooled by liquid helium from 1.5 K [12]. Three decades later, in 1967, Brown constructed a reciprocating magnetic refrigerator to demonstrate the feasibility of room-temperature magnetic refrigeration using Gd metal[13], achieving a temperature span of 47 K after 50 cycles. After this period, MCE was not extensively investigated until 3 decades later when Pecharsky and Gschneider discovered the giant magnetocaloric effect (GMCE) in Gd5Si2Ge2 in 1977[7]. For this alloy, the MCE was enhanced due to a structural transformation that comes with the magnetic transition. Following this discovery, the number of publications on the magnetocaloric effect and magnetic refrigeration was steadily increasing with some recent stabilization trend (Figure 1.1). Figure 1.1: Publications about magnetocaloric effect over past decades (SCOPUS). 7 Chapter 1 Enhanced properties of magnetocaloric materials achieved over the last decades have increased the interest in using them for solid-state refrigeration instead of a gas-expansion-compression technology of a conventional refrigeration. The replacement of traditional technology by solid state refrigeration, particularly, by MCE cooling, offers several advantages: ❖ Environmentally-friendliness since it does not use ozone-depleting gases like (CFC’s, HCFC’s) ❖ Higher efficiency (60% of Carnot efficiency compared to 40% for conventional refrigeration), requiring less energy and resulting in reduced CO2 emissions ❖ Lower costs ❖ Minimal maintenance and reduced noise Figure 1.2: Visualization of magnetocaloric cooling in a single cycle versus its analogue for the vapour compression system. Figure 1.2 illustrates the fundamental working mechanism of magnetic refrigeration in comparison to its vapour expansion-compression refrigerator counterpart. In the case of conventional refrigeration, a gas undergoes adiabatic compression and decompression, resulting in heat generation during 8 Introduction compressing and heat dissipation during adiabatic decompression. Analogously, in solid-state refrigeration, the cycle starts with the material at thermal equilibrium, with magnetic moments randomly oriented.When a magnetic field is applied under an adiabatic conditions, the magnetic dipoles align with the magnetic field, causing the sample to experience a temperature increase due to a decrease in its magnetic entropy. After the increased heat dissipates through a radiator, the material returns to its initial temperature. When the magnetic field is removed adiabatically, the material experiences a temperature decrease, reaching a lower temperature than the initial one. This stage can be used for refrigeration. After this, the material thermally stabilizes and the cycle starts again. Both conventional and inverse magnetocaloric effects are characterized by the isothermal entropy change, ΔSm(T, H), and/or the adiabatic temperature change, ΔTad(T, H), when a magnetic field is applied or removed under isothermal or adiabatic conditions, respectively[14][15]. In the most easiest way, the magnetic field induced entropy change can be estimated using a thermodynamic Maxwell relationship[16], although in the case of the first-order magnetostructural transformations some additional considerations should be taken into account [16][17]. Commonly, thermomagnetic curves measured under the iso-field conditions are used to calculate ΔSm(T, H) through numerical approximation of the Maxwell relationship: ∆𝑆𝑚(𝑇,𝐻)= 𝑆𝑚(𝑇,𝐻)−𝑆𝑚(𝑇,0)=∫(𝜕𝑀(𝑇,𝐻′) 𝜕𝑇 )𝑑𝐻′ 𝐻 0 (3) 1.2.1.1. Conventional magnetocaloric effect In terms of energy, the magnetocaloric effect is characterized as the change of magnetic entropy in an isothermal process or as the change of temperature under adiabatic conditions upon the applications of an external magnetic field (∆𝐻). The physical process of the MCE is related to the coupling 9 Chapter 1 between the external magnetic field and the magnetic moments within the material. First, we will explain the MCE mechanism with a pictorial description and then we will provide a more formal thermodynamic interpretation. Let us assume a magnetic material is inside and outside a magnetic field (see Figure 1.3). When no magnetic field is applied, the magnetic moments are randomly oriented, giving rise to some entropy, which we will refer to as magnetic entropy, SM. If we adiabatically apply a magnetic field to the material (Fig 1.3 left), the magnetic moments will become oriented with the field (let us assume fully oriented, for simplicity) and the magnetic entropy becomes zero (SM = 0), resulting in a negative change in magnetic entropy (SM < 0). Since the process is adiabatic, the total entropy change is zero (∆S = 0). The total entropy change includes lattice (SL), electronic (SE) and magnetic (SM) contributions so: ∆𝑆=∆𝑆𝐿+∆𝑆𝑀+∆𝑆𝐸=0⇒∆𝑆𝐿+∆𝑆𝐸=−∆𝑆𝑀>0 (4) This positive change in electron and lattice entropy implies an increase in temperature of the material. When the magnetic field is adiabatically removed, the magnetic moments becomes disoriented, resulting in the cooling of the material. For MCE measurement, the adiabatic temperature change (∆𝑇𝑎𝑑) achieved in this process of magnetization/demagnetization is used as a parameter to evaluate the material’s cooling efficiency. Figure 1.3: Schematic representation of application of an external magnetic field under adiabatic (left) and isothermal (right) conditions. 10 Introduction Apart from the adiabatic temperature change, the MCE can also be represented by the isothermal entropy change, ∆𝑆𝑖𝑠𝑜. Figure 1.3 right shows the magnetization process under isothermal conditions. In this case, magnetic entropy reaches zero after the alignment of the magnetic moments with the magnetic field, so the magnetic entropy change is negative (∆𝑆𝑀<0). This magnetic entropy, associated with magnetic moments, is “released” to the environment because the temperature remains constant, making it equals to ∆𝑆𝑖𝑠𝑜 (assuming there is no coupling between the three contributions to entropy stated above). Unlike adiabatic temperature change, ∆𝑆𝑖𝑠𝑜 is relatively easy to measure and is the most commonly reported parameter in MCE evaluations. From these considerations, we can deduce that the largest entropy change and thus the largest adiabatic temperature change will be obtained when magnetic moments become fully disoriented at 𝐻=0 and fully aligned at 𝐻≠ 0. The materials that can satisfy these conditions at ambient temperatures and achievable magnetic fields in practice are ferromagnets. For these materials, the magnetic moments naturally align themselves at the Curie temperature and ferromagnetic ordering can be achieved under relatively small magnetic fields. In the case of paramagnets, spontaneous ferromagnetic ordering does not occur and large magnetic fields are required to induce such order, making ferromagnets the only ones of practical interest. Figure 1.4: Entropy versus temperature behaviour of a magnetic material under no magnetic field and under a magnetic field for a first-order magnetic ordering (left) and a second-order magnetic ordering (right). 11 Chapter 1 The magnetic entropy change depends on both magnetic field and temperature. The temperature evolution at two different magnetic fields is graphically depicted in Figure 1.4, where we can distinguish two processes: ❖ Isothermal magnetization change (vertical line). The magnetic field is applied isothermally, reducing the entropy by ∆𝑆𝑖𝑠𝑜, while lattice and electronic contributions remain constant. Magnetic entropy change is given by: ∆𝑆𝑀(𝑇)𝑇,∆𝐻,𝑃 =[𝑆𝑀(𝑇)𝐻1−𝑆𝑀(𝑇)𝐻0]𝑇,𝑃 =[𝑆1(𝑇)𝐻1−𝑆0(𝑇)𝐻0]𝑇,𝑃 (5), where 𝑆𝑀 is the magnetic entropy and 𝑆0,𝑆1 are the total entropies at different magnetic fields. ❖ Adiabatic magnetization change (horizontal line): Applying the magnetic field under adiabatic conditions leads to the ordering of magnetic moments. The magnetic entropy decreases whereas the electronic and lattice entropies increase to maintain the total entropy constant. This results in an increase of the temperature in the material by a value ∆𝑇𝑎𝑑, expressed as: ∆𝑇𝑎𝑑(𝑇)𝑇,∆𝐻,𝑃 =[𝑇1(𝑆)𝐻1−𝑇0(𝑆)𝐻0]𝑆,𝑃 (6), where 𝑇0 and 𝑇1 are the temperatures in magnetic fields 𝐻0 and 𝐻1, respectively. The magnetocaloric effect can be treated from a more formal thermodynamic approach. The entropy, which is the first derivative of the Gibbs free energy (𝑆=−(𝜕𝐺/𝜕𝑇)𝑃), is a continuous function of temperature but changes its slope at the transition temperature. While entropy cannot be measured directly, it can be calculated from an experimentally measurable quantity, namely, the heat capacity, 𝐶(𝑇)𝐻,𝑃: 𝑆(𝑇)𝐻,𝑃 =∫𝐶(𝑇)𝐻,𝑃 𝑇𝑑𝑇 𝑇 0 (7) 12 Introduction The total entropy of the material as a function of temperature is shown in Figure 1.4 left for a first-order magnetic transition. In this case, there is a discontinuous change of entropy at transition temperature T1. The jump in the entropy comes from the enthalpy of the transition, ∆𝐸, and the entropy increase is ∆𝐸/𝑇. The isothermal entropy change at temperature T (𝑇𝑡1 <𝑇<𝑇𝑡2) is: ∆𝑆𝑖𝑠𝑜(𝑇)∆𝐻,𝑃 =∫𝐶𝑙(𝑇)𝐻2,𝑃 𝑇𝑑𝑇 𝑇 0−∫𝐶𝑙(𝑇)𝐻1,𝑃 𝑇𝑑𝑇 𝑇𝑡1 0−∫𝐶ℎ(𝑇)𝐻1,𝑃 𝑇𝑑𝑇 𝑇 𝑇𝑡1 −∆𝐸𝐻1 𝑇𝑡1 (8) where 𝐶𝑙 and 𝐶ℎ are the heat capacities at low- (below transition temperature) and high-temperature (above transition temperature) phases. If we assume that 𝐶𝑙(𝑇)𝐻,𝑃 ≈𝐶ℎ(𝑇)𝐻,𝑃 as observed in Gd5Si2Ge2 for the same magnetic field[18], we get: ∆𝑆𝑖𝑠𝑜(𝑇)∆𝐻,𝑃 =∫𝐶(𝑇)𝐻2,𝑃 −𝐶(𝑇)𝐻1,𝑃 𝑇𝑑𝑇 𝑇 0−∆𝐸𝐻1 𝑇1 (9) From this equation it can be deduced that the larger difference between the heat capacities in two fields gives a larger entropy change. The last term in this equation, called the structural entropy change, ∆𝑆𝑠𝑡𝑟, is an additional contribution associated with the first-order phase transition and the abrupt atomic rearrangement that happens in this class of transitions. The relation between the entropy change and the adiabatic temperature change is given by[15]: ∆𝑇𝑎𝑑(𝑇)∆𝐻,𝑃 ≅ − 𝑇 𝐶(𝑇0)𝐻2,𝑃∆𝑆𝑖𝑠𝑜(𝑇)∆𝐻,𝑃 (10) where 𝑇0∈[ 𝑇,𝑇+∆𝑇𝑎𝑑(𝑇)∆𝐻,𝑃 ] and is unknown. In any case, from the above equation we see that ∆𝑇𝑎𝑑 is directly proportional to entropy change and inversely proportional to the heat capacity, so it will be lower for materials with a high heat capacity. 13 Chapter 1 When the material undergoes a second-order magnetic transition (Figure 1.4 right), the entropy is a continuous function of temperature and changes its slope at the transition temperature. In this case, the isothermal entropy change is given by: ∆𝑆𝑖𝑠𝑜(𝑇)∆𝐻,𝑃 =∫𝐶(𝑇)𝐻2,𝑃 −𝐶(𝑇)𝐻1,𝑃 𝑇𝑑𝑇 𝑇 0 (11) From this equation, we can see that the entropy change will be larger when the difference between the heat capacities in two fields is large. The expression for the entropy change, Eqs.(9,11), are the theoretical ones. In reality, only 6090% of this can be achieved. Moreover, only a small fraction of this entropy (less than 30%) is used in a magnetocaloric process, so experimental values of ∆𝑆𝑚𝑎𝑔 are significantly smaller than the theoretical ones. 1.2.1.2. Inverse magnetocaloric effect The inverse magnetocaloric effect is the opposite phenomenon to the conventional one. In this case, the magnetic entropy change is positive when the adiabatic temperature change is negative, and vice versa. The sample cools down when a magnetic field is applied. This effect is found in materials that undergo a reverse martensitic transformation, from paramagnetic/antiferromagnetic martensite to ferromagnetic austenite. It is observed in NiMnX (X=Sn, In, Sb) Heusler-type alloys and has been extensively studied[19][20][21]. 1.2.1.3. Refrigeration capacity MCE is also characterized by a refrigerant capacity (RC) which is defined as the amount of heat transferred between hot and cold reservoirs in a single refrigeration cycle. There are three different ways for calculating the RC, as reported in literature: 14 Introduction (1) Relative cooling power (RCP), determined by the product |∆𝑆𝑀( 𝐻)|𝑚𝑎𝑥 × 𝛿𝑇𝐹𝑊𝐻𝑀(𝐻) [19] (2) Introduced by Gschneidner, where the area under the peak of ∆𝑆𝑀 within 𝛿𝑇𝐹𝑊𝐻𝑀 temperature range is calculated[22]: 𝑅𝐶=∫|∆𝑆𝑀(𝑇,𝐻)|𝑑𝑇 𝑇ℎ𝑜𝑡 𝑇𝑐𝑜𝑙𝑑 (12) (3) By Wood and Potter, where the area of the largest rectangle of ∆𝑆𝑀 is estimated[23] Figure 1.5: Schematic representation of the RC obtained from the magnetic entropy change curve. The RCP blue rectangle has the same width as the Gschneidner area’s maximum width. These three calculations are shown graphically in Figure 1.5. Since both the entropy change and adiabatic temperature change are proportional to the derivative of magnetization with respect to temperature, |𝜕𝑀/𝜕𝑇|, the greater is the variation of magnetization with temperature, the higher values of MCE are achieved. This is expected for both firstand second-order phase transitions. The MCE occurs at the Curie temperature in second-order phase transitions, i.e., transition from paramagnetic to ferromagnetic ordering. 15 Chapter 1 1.2.2. Elastocaloric effect The elastocaloric effect (eCE) is the thermal response to an external mechanical stress. This external stimulus induces a phase transformation, resulting in an entropy change and, consequently, a temperature change in the material. Elastocaloric materials reported for cooling mainly include NiTibased, Fe-based, Cu-based and ferromagnetic SMAs. Rubber, as an example, increases its temperature when rapidly stretched, as first reported by Gough in 1805, who found out that rubber heats slightly when stretched rapidly[24]. Soon after, Thomson proposed a thermodynamic interpretation [25], and Joule discovered several elastocaloric materials[26]. However, due to the weak caloric effects of common metals and polymers, the elastocaloric effect barely received attention in the next 100 years. In 1980, Rodriguez and Brown, studying the martensitic transformation, occasionally found a significant eCE in Cu69.6Al27.7Ni2.7 SMA [27]. In 2004, Quarini and Prince reported large temperature variations of 16K and -14K in the NiTi alloy subjected to a loadingholding-unloading protocol and originally proposed the concept of solid-state cooling[28]. As already mentioned, the elastocaloric effect arises from the absorption or release of latent heat during the martensitic transformation that occurs during cyclic loading and unloading. Figure 1.6 shows one cycle of loading and unloading for a superelastic NiTi alloy. This Brayton cycle consists of four steps: (i) Adiabatic loading (1 → 2): in this step, an exothermic martensitic transformation from highly-ordered cubic austenitic phase to lowsymmetry monoclinic martensite phase occurs, resulting in the heating of the material. (ii) Heating (2 → 3): The applied stress or strain is maintained at a constant value while the material releases the heat acquired in the first 22 Introduction 1.4.1. NiMn-based Heusler alloys Ni-Mn-based Heusler-type MetaMagnetic Shape Memory Alloys (MMSMAs), a set of alloys that present the Magnetocaloric Effect (MCE), are of great interest due to their strong potential for a solid-state refrigeration. They undergo a first-order martensitic transformation from a high-symmetry ferromagnetic austenitic phase at high temperature to a low-symmetry weak magnetic or antiferromagnetic martensitic phase at lower temperature, whereby exhibiting a large change of the magnetization[37]. This magnetostructural transformation leads to the so-called magnetic shape memory effect[38]. In addition, the application of an external magnetic field can shift the MT because of the strong magnetostructural coupling[39][40][41], giving rise to the peculiar giant MCE phenomena, making these materials a promising candidates for efficient solid-state refrigeration applications. The phase stability as well as structural and magnetic properties of Ni-MnX (X=In, Sn, Sb, Ga,…) Heusler alloys have been widely studied both theoretically and experimentally in the literature, showing that it is possible to manipulate, in the predictable way, the magnetic exchange interactions in both low-temperature martensite and high-temperature austenite, as well as tune other multifunctional properties, such as shape memory effect or superelasticity, by the doping with many other elements, such as Co, Cu, Fe, Cd, W etc. [42][43][44][45][46][47][48]. Ni-Mn-based Heusler MMSMAs are very attractive candidates for magnetocaloric applications owing to the availability of the raw materials, non-toxicity, easiness-to-be-prepared and a large magnetocaloric effect under a magnetic field of existing permanent magnets[49][50]. Ni-Mn-Sn Heusler alloys family represents one of the intensively studied MMSMAs. This stoichiometric compound, Ni2MnSn, has a L21 –ordered cubic structure with four interpenetrating face centered cubic (fcc) sublattices[51] (Figure 1.10). In an ideal ordered case, the (0,0,0) and (12 ⁄ ,12 ⁄ ,12 ⁄ ) sites are 23 Chapter 1 occupied by Ni atoms, leaving the remaining (14 ⁄ ,14 ⁄ ,14 ⁄ ) and (34 ⁄ ,34 ⁄ ,34 ⁄ ) sites being occupied by Sn and Mn atoms. On the other hand, in off-stoichiometric Ni-Mn-Sn Heusler alloys, the excess of Mn atoms occupy the partially vacant Sn sites. Indeed, some percentage of Ni, Mn, Sn atoms may be distributed randomly, forming some degree of disorder present in the crystal structure which can be largely removed by annealing these materials at high temperatures[52][53]. Figure 1.10: B2 vs L21 structures for austenitic phase. Upon cooling from the melt, Ni-Mn-X (X= Sn, Sb, In, Ga) alloys crystallize first in entirely disordered A2 structure, then exhibit a partially ordered B2 structure, where Ni atoms occupy the corner positions and Mn and Z atoms are randomly located in body-centered (bcc) positions. Further cooling produces the disorder-order transition where the crystal structure of alloys transforms from B2 to L21, in which Ni atoms are located in the corners of the structure, whereas Mn and Z atoms are at alternate body sites. The martensite transformation temperature depends on the atomic order, which can be modified by the chemical composition change and/or by the processing of the alloy. The Ni-Mn-X alloys exhibit a martensitic transformation into martensites with a non-modulated or modulated crystal lattices. When the concentration of X atoms is low this family of alloys transforms into a non-modulated tetragonal 24 Introduction L10 –ordered martensite [54]. Figure 1.11 shows the crystallographic lattice relations between L21 austenite and L10 martensite. Martensite may have modulated structures apart from L10 structure (especially for alloys with a high Z concentration). The most common modulated structures are the following ones: (1) Four-layered 4O structures[55] (2) Five-layered 10M structure [56][57] (3) Seven-layered 14M structure[56] These modulated structures are formed by shearing of the (110) planes along the [11 0] crystallographic direction. The fabrication process and chemical composition significantly influence the crystal structure of martensite and the equilibrium martensitic transformation temperature, Tm0. The off-stoichiometric Ni-deficient Ni-Mn-Sn Heusler alloys can be doped with Cobalt, where Cobalt goes to the Ni sites. These Ni(Co)-Mn-Sn Heusler alloys are of great interest since they reportedly have MT near room temperature[58]. Previous works have investigated the crystal structures of martensite phases in these alloys by means of X-ray diffraction (XRD). Umetsu et al. showed that Co-doped Ni-Mn-Sn alloy has an L21 structure (Figure 1.11) at room temperature[59], with space group Fm-3m (Cu2MnAl prototype). This structure is confirmed in the present work using Mn-rich Ni40Mn42.5Co8Sn9.5 alloy at room temperature [60]. Umetsu et al. concluded that in the Ni-Mn-Sn alloy, the Mn moments on 4a and 4b sites are antiferromagnetically coupled, whereas in the Co-doped alloy they are ferromagnetically coupled, attributing the ferromagnetic enhancement to the change in magnetic structure by Co substitution. On the other hand, melt-spinning is a useful technique for obtaining ready-shaped magnetocaloric materials with a high surface/volume ratio suitable for their implementation in active magnetic regenerators. Thanks to the 25 Chapter 1 high cooling rate of melt-spinning (about 106 K/s) it can modify the physical characteristics of the alloys. Such high rates induce the stabilization of the high temperature austenite phase in Heuler-type MMSMAs with its corresponding B2 ordered structure instead of the highly ordered L21 structure. As far as MCE and other functionalities of Ni-Mn-based Heusler alloys are closely related to the crystal structures of constituent phases, a deep understanding of crystal structure is required for exploiting all these functionalities towards property optimization. Figure 1.11: Crystallographic relations between L21 austenite and non-modulated L10 martensite. For NiMnSn stoichiometric alloys, X are Ni atoms, Y are Mn atoms and Z are Sn atoms. The Heusler-type MMSMAs are very sensitive to composition in terms of phase transformation temperature[61][62], Curie temperature[63], saturation magnetization[64], which, in turn, are direct consequence of a crystal structure variation[65]. Also, post-annealing and cooling rates have a strong influence on the magnetic properties[66], as phase changes occur due to atomic ordering. Material processing may degrade the magnetic properties of these compounds, but these can be reversed by performing heat treatments[60]. 1.4.1.1. Effects of doping (a) Co addition The modification of the Ni/Mn/Sn ratio in Heusler NiMn-based alloys influences their magnetocaloric properties and transformation characteristic 26 Introduction temperatures. Nevertheless it is difficult to shift the martensitic transformation towards room temperature with just varying that ratio, while simultaneously having a large magnetocaloric effect at low fields. Adding Co to Ni-Mn-Sn alloys allows to obtain suitable materials for magnetic cooling due to the strong influence of Co on magnetic and structural properties. Ni-Co-Mn-Sn system has been widely studied by several groups. Krenke et al. studied the influence of substituting Co for Ni in Ni50-xCoxMn37Sn13 alloys on the magnetocaloric effect[67]. They found that introducing Co led to a decrease in Ms and ΔS but the thermal hysteresis associated with the transition became narrower with Co content from 1 to 3 at.%. Similarly, Cong et al. performed a systematic study of the Ni50-xCoxMn39Sn11 alloys and obtained a phase diagram relating the chemical composition to temperatures of phase transitions and magnetic behaviour[68]. puta Figure 1.12 shows the relationship between transformation temperatures and the Co content, ranging from 0 to 19 at.%. The martensitic transformation temperature, defined in this study as (Tms+Tmf+Tas+Taf)/4, shows a roughly linear negative dependence with increasing Co content until 7 at.% of Co and strong negative exponential decrease for more than 7 at.% of Co content. The linear slow decrease of TM is ascribed to the change of e/a ratio. The substitution of Ni (10 valence electron) by Co (9 valence electron) decreases the e/a and, as consequence, leads to a decrease of TM. For higher Co content, the rapid decrease may be due to atomic order, precipitation of other phases, etc. In particular, the formation of the secondary 𝛾 phase was observed that has a different composition than the matrix. The austenite Curie temperature increases with increasing Co content and seems to be correlated to the strengthening of the ferromagnetic exchange interactions with Co doping. Based on the diagram, it can be noted that the most interesting range of compositions, from a practical point of view, is between 5 and 8 at.% of Co since in this region MT occurs with large magnetization change that benefits the magnetocaloric effect. The addition of Co also affects the crystal structure 27 Chapter 1 with the formation of modulation from 4O to 10M and then to 14M with an increase of Co content. Figure 1.12: Phase diagram for Ni-Co-Mn-Sn system. The evolution of the characteristic transformation temperatures versus the Co content is displayed [68]. (b) Cu addition Cu addition to the Ni-Mn-Sn Heusler alloys also enhances the magnetocaloric properties and influences the crystal structure and phase transformation temperatures. Das et al. investigated the alloys system of Ni44xZxMn43Sn11 (Z = Co, Cu) and the effect of Co versus Cu doping for magnetocaloric properties as well as for transformation temperatures. Cu doping led to a decrease in both martensitic and magnetic transformation temperatures. Substituting Cu for Ni increased ΔSM significantly, having an even stronger effect than the addition of Co[68]. The decrease of Tms with Cu substitution for Ni is opposite to what would be expected by the e/a rule. In this case, another factor such as the unit cell volume should be considered. On the other hand, the replacement of Mn by Cu in Ni43Mn46-xCuxSn11 increases 28 Introduction both magnetic and martensitic transition temperatures. In this case, ΔSM increases with the addition of Cu[69]. The magnetocaloric properties are improved due to a larger magnetization change at martensitic transformation from austenitic to martensitic phases caused by Cu addition. It was demonstrated that Cu enhances ferromagnetic exchange interaction in austenite resulted in the mentioned greater magnetization change at MT[70]. 1.4.1.2. Effect of fabrication methods The fabrication method has a significant effect on the properties of the resulting alloy. For alloy synthesis, the most popular methods are induction casting and arc-melting. In the laboratory conditions, the former technique offers the possibility to prepare alloys in large quantities (several tens of grams), whereas the latter one should not exceed a few grams. The reason is that induction casting uses large crucible in which a large quantity of alloy fits and that can be completely and simultaneously melted, giving rise to a rather homogeneous liquid alloy that will maintain its homogeneity when casted at a high quenching rate. In this method, post annealing is required for improving a homogeneity. Another drawback of this technique is that the alloy is susceptible for reaction with the alumina crucible if overheated in a liquid state or keeping liquid too long. On the other hand, arc melting furnace generally provides homogeneous alloy as long as its mass is less than 10 grams since otherwise the arc cannot melt the entire constituents simultaneously, preventing it from becoming a fully homogeneous ingot. Flipping and re-melting the ingot a few times assure the alloy homogeneity. Melt-spinning is a technique that produces metallic ribbons or flakes which could be a single-phase without chemical segregation. The high cooling rate in normally prevents the formation of unexpected phases[71]. This technique allows to shorten or even avoid a long annealing, leading to lower fabrication cost and the fabrication time. In addition, melt-spinning process 29 Chapter 1 influences atomic order, which affects the magnetic properties and martensitic transition temperatures. Generally, MMSMAs ribbons have a lower MT than the bulk alloys, and this is attributed to grain refinement and internal stresses formed during fabrication [72]. 1.4.1.3. Heat treatments The functional properties of a material can be improved by performing specific heat treatments. On the other hand, material processing like meltspinning, grinding process for powder obtaining, or any post processing may degrade its magnetic properties. This degradation can be partially or completely reverted by performing specific heat treatments to the material. Another property that can be enhanced is the thermal hysteresis reduction since heat treatments have influence on the microstructure, phase transitions, and magnetic properties of these materials. Heat treatments affect in various manners to the material: 1) Phase Transition Stabilization: Heusler magnetocaloric materials typically undergo phase transitions when exposed to varying magnetic fields. These phase transitions are essential for the magnetocaloric effect, but they can introduce hysteresis. Heat treatments can stabilize these phase transitions, making them more reproducible and reducing hysteresis. 2) Microstructural changes: Heat treatments can induce changes in the microstructure of the material. For example, they can promote the growth of specific crystal grains or the elimination of defects. These microstructural changes can result in a more uniform and well-defined response of the material to changes in temperature and magnetic field. 30 Introduction 3) Homogenization: Heat treatments can help in the homogenization of the material composition. Inhomogeneities or compositional variations within the material can lead to irregular phase transitions and hysteresis. Heat treatments can even out these variations and improve the material's consistency. 4) Magnetic property optimization: The magnetic properties of Heusler magnetocaloric materials are closely linked to their phase transitions. Heat treatments can help optimize these magnetic properties, such as the Curie temperature and the saturation magnetization. 5) Stress Relief: During the fabrication process, magnetocaloric materials may experience stress or strain, which can affect their magnetic and thermal behaviour. Heat treatments can relieve these internal stresses and restore the material to a more stable and stress-free state. 6) Defect Annealing: Any defects or dislocations in the crystal lattice of the material can disrupt its magnetic and thermal behaviour. Heat treatments can facilitate defect annealing, reducing these disruptions and promoting more predictable phase transitions. Overall, heat treatments are a valuable tool in optimizing the properties of Heusler magnetocaloric materials. They help achieve more controlled and predictable phase transitions, leading to a reduction in thermal hysteresis and, ultimately, enhancing the efficiency and reliability of these materials for cooling and refrigeration applications. 31 Chapter 1 1.5. APPLIED ASPECTS OF MAGNETOCALORIC MATERIALS puta The main potential application of magnetocaloric materials is for magnetic refrigeration. Both first-order magnetocaloric (FOMT) and second-order magnetocaloric (SOMT) materials are generally used for magnetic refrigeration and heat pumping[73]. It is clear that a magnetocaloric material exhibiting a large magnetocaloric effect is desirable for refrigeration. One of the parameters for measuring the utility of a magnetocaloric material is its magnetic field – induced isothermal entropy change, a quantity that is usually reported in every study devoted to a magnetocaloric material. However, this is not only parameter needed to be taken into account since it does not provide information about the usability of the material for a refrigeration device, because heat is not transferred isothermally. The main parameter that must be taken into account is the magnetic field –induced adiabatic temperature change since it provides a basis for establishing the capacity of magnetocaloric material to create the temperature gradient. The largest possible adiabatic temperature change is required to overcome the irreversible heat losses because of the irreversibility of the heat transfer between the material and the heat transferring fluid. puta Therefore, it is desirable to have magnetocaloric materials with the largest adiabatic temperature change per magnetic field unit, taking into consideration environmental, geopolitical and resources issues associated with practical applications. Gd-based materials are considered critical in this regard [74], while the chemical elements in NiMn-based Heusler compounds are in a common use, which is one of the reasons for choosing these materials for this thesis. 1.5.1. Active Magnetic Regenerators The temperature span required for magnetocaloric refrigeration is greater than the one a magnetocaloric material can provide by itself, although it is enough for low temperature adiabatic demagnetization refrigeration due to the 38 Introduction 1.8. OBJECTIVES A global objective of the present work is to develop practically new functional material such as powdered high efficient Heusler-type MMSMA exhibiting MT and MCE characteristics similar to a bulk homolog and being suitable for additive manufacturing of a heat exchanger prototype. We will pursue the next less scoped objectives: 1. Exploring transformation behaviour of the representatives of the main families of Heusler-type NiMn-based MMSMAs in order to select a prospective alloy which will serve, at the end, as prototype MCE powder component of printing ink. 2. Preparation and all-round study of MMSMA ribbons as the precursors for fabrication of the MCE functional powder 3. All-round investigations of MMSMAs powders prepared from ribbons 4. Elucidating conditions needed to create metallic printable inks 5. 2D and 3D printing of MCE prototype devices using a developed MMSMA/polymer inks and unveiling their characteristics. 39 Chapter 1 1.9. REFERENCES [1] International Energy Agency (IEA), The Future of Cooling Opportunities for energyefficient air conditioning, (2018) 92. https://iea.blob.core.windows.net/assets/0bb45525-277f-4c9c-8d0c9c0cb5e7d525/The_Future_of_Cooling.pdf. [2] O. Sari, M. Balli, From conventional to magnetic refrigerator technology, Int. J. Refrig. 37 (2014) 8–15. https://doi.org/10.1016/j.ijrefrig.2013.09.027. [3] Zimm et al., DESCRIPTION AND PERFORMANCE OF A NEAR-ROOM TEMPERATURE MAGNETIC REFRIGERATOR, 43 (n.d.) 1759–1760. [4] C.R.H. Bahl, T.F. Petersen, N. Pryds, A. Smith, A versatile magnetic refrigeration test device, Rev. Sci. Instrum. 79 (2008). https://doi.org/10.1063/1.2981692. [5] Y. Mozharivskyj, Magnetocaloric Effect and Magnetocaloric Materials, Elsevier Inc., 2016. https://doi.org/10.1016/b978-0-12-409547-2.11643-9. [6] L. Mañosa, A. Planes, M. Acet, Advanced materials for solid-state refrigeration, J. Mater. Chem. A. 1 (2013) 4925–4936. https://doi.org/10.1039/c3ta01289a. [7] V.K. Pecharsky, J. Gschneidner K. A., Giant Magnetocaloric Effect in Gd5Si2Ge2, Phys. Rev. Lett. 78 (1997) 4494–4497. https://doi.org/10.1103/PhysRevLett.78.4494. [8] Thomson, W. In Cyclopedia of the Physical Sciences; Nichol, J. P., Ed.; 2nd ed.; Richard Green and Company: London; p 838., (1860). [9] W. Thomson, On the thermoelastic, thermomagnetic, and pyroelectric properties of matter, London, Edinburgh, Dublin Philos. Mag. J. Sci. 5 (1878) 4–27. https://doi.org/10.1080/14786447808639378. [10] E. Warburg, ., E. Ann. Phys. Chem. (1881) 141. https://doi.org/10.1093/benz/9780199773787.article.b00194542. [11] A. Smith, Who discovered the magnetocaloric effect? Warburg, Weiss, and the connection between magnetism and heat, Eur. Phys. J. H. 38 (2013) 507–517. https://doi.org/10.1140/epjh/e2013-40001-9. [12] W.F. Giauque, D.P. MacDougall, Attainment of temperatures below 1°absolute by demagnetization of Gd2(SO4)38H2O [12], Phys. Rev. 43 (1933) 768. https://doi.org/10.1103/PhysRev.43.768. [13] G. V. Brown, Magnetic heat pumping near room temperature, J. Appl. Phys. 47 (1976) 3673–3680. https://doi.org/10.1063/1.323176. [14] T. Krenke, E. Duman, M. Acet, E.F. Wassermann, X. Moya, L. Manosa, A. Planes, Inverse magnetocaloric effect in ferromagnetic Ni-Mn-Sn alloys, Nat. Mater. 4 (2005) 450–454. https://doi.org/10.1038/nmat1395. [15] V.K. Pecharsky, K.A. Gschneidner, A.O. Pecharsky, A.M. Tishin, Thermodynamics of the magnetocaloric effect, Phys. Rev. B. 64 (2001) 144406. https://doi.org/10.1103/PhysRevB.64.144406. [16] V.K. Pecharsky, K.A. Gschneidner, Magnetocaloric effect from indirect measurements: Magnetization and heat capacity, J. Appl. Phys. 86 (1999) 565–575. 40 Introduction https://doi.org/10.1063/1.370767. [17] V.A. Chernenko, V.A. L’vov, E. Cesari, J.M. Barandiaran, Fundamentals of magnetocaloric effect in magnetic shape memory alloys, 1st ed., Elsevier B.V., 2019. https://doi.org/10.1016/bs.hmm.2019.03.001. [18] V.K. Pecharsky, K.A. Gschneidner, Heat capacity near first order phase transitions and the magnetocaloric effect: An analysis of the errors, and a case study of Gd5(Si2Ge2) and Dy, J. Appl. Phys. 86 (1999) 6315–6321. https://doi.org/10.1063/1.371734. [19] K.A. Gschneidner, V.K. Pecharsky, Magnetocaloric materials, (2000) 387–429. [20] A. You, M.A.Y. Be, I. In, Hysteresis effects in the inverse magnetocaloric effect in martensitic Ni-MnIn and Ni-Mn-Sn, 073914 (2015) 5–10. [21] J.A. Phys, T. Krenke, E. Duman, M. Acet, Effect of Co and Fe on the inverse magnetocaloric properties of Ni-Mn-Sn, 033903 (2015) 1–6. https://doi.org/10.1063/1.2761853. [22] K.A. Gschneider, Recent develompents in magnetic refrigeration, 317 (1999) 69–76. https://doi.org/10.4028/www.scientific.net/MSF.315-317.69. [23] D. Salazar-Jaramillo, P. Álvarez-Alonso, P. Lázpita, J.L. Sánchez Llamazares, P. Gorría, J.A. Blanco, V.A. Chernenko, Magnetocaloric Effect in Specially Designed Materials, Elsevier Inc., 2018. https://doi.org/10.1016/B978-0-12-813904-2.000073. [24] J. Gough, No Title, Mem. Lit. Phil. Soc. Manchester. 1 (2nd Ser (1805) 288. [25] W. Thomson, On the thermoelastic and thermomagnetic properties of matter, in: Quart. J. Pure Appl. Math., 1857: pp. 57–77. [26] J.P. Joule, On some thermo-dynamic properties of solids, in: Philos. Trans. R. Soc. L., 1859: pp. 91–131. [27] C. Rodriguez, L.C. Brown, The thermal effect due to stress-induced martensite formation in Β-CuAlNi single crystals, Metall. Mater. Trans. A. 11 (1980) 147–150. https://doi.org/10.1007/BF02700450. [28] J. Quarini, A. Prince, Solid state refrigeration: Cooling and refrigeration using crystalline phase changes in metal alloys, Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 218 (2004) 1175–1179. https://doi.org/10.1243/0954406042369062. [29] J. Van Humbeeck, Shape memory alloys: A material and a technology, Adv. Eng. Mater. 3 (2001) 837–850. https://doi.org/10.1002/15272648(200111)3:11<837::AID-ADEM837>3.0.CO;2-0. [30] P.J. Webster, Heusler alloys, Contemp. Phys. 10 (1969) 559–577. https://doi.org/10.1080/00107516908204800. [31] T. Graf, C. Felser, S.S.P. Parkin, Simple rules for the understanding of Heusler compounds, Prog. Solid State Chem. 39 (2011) 1–50. https://doi.org/10.1016/j.progsolidstchem.2011.02.001. [32] N. V. Uvarov, Y. V. Kudryavtsev, A.F. Kravets, A.Y. Vovk, R.P. Borges, M. Godinho, V. Korenivski, Electronic structure, optical and magnetic properties of Co 2FeGe Heusler alloy films, J. Appl. Phys. 112 (2012). https://doi.org/10.1063/1.4752870. 41 Chapter 1 [33] M. Khan, I. Dubenko, S. Stadler, N. Ali, Magnetostructural phase transitions in Ni50Mn 25+xSb25-x Heusler alloys, J. Phys. Condens. Matter. 20 (2008). https://doi.org/10.1088/0953-8984/20/23/235204. [34] W.J. Buehler, J. V. Gilfrich, R.C. Wiley, Effect of Low-Temperature Phase Changes on the Mechanical Properties of Alloys near Composition TiNi, J. Appl. Phys. 34 (1963) 1475–1477. https://doi.org/10.1063/1.1729603. [35] M. Kelansth, A. Dunbar, ELrgonne Rational daboratoq!, (1962). [36] T. Gottschall, K.P. Skokov, M. Fries, A. Taubel, I. Radulov, F. Scheibel, D. Benke, S. Riegg, O. Gutfleisch, Making a Cool Choice: The Materials Library of Magnetic Refrigeration, Adv. Energy Mater. 9 (2019). https://doi.org/10.1002/aenm.201901322. [37] Y. Wang, D. Salas, T.C. Duong, B. Medasani, A. Talapatra, Y. Ren, Y.I. Chumlyakov, I. Karaman, R. Arróyave, On the fast kinetics of B2–L21 ordering in Ni-Co-Mn-In metamagnetic shape memory alloys, J. Alloys Compd. 781 (2019) 479–489. https://doi.org/10.1016/j.jallcom.2018.12.034. [38] R. Kainuma, Y. Imano, W. Ito, Y. Sutou, H. Morito, S. Okamoto, O. Kitakami, K. Oikawa, a Fujita, T. Kanomata, K. Ishida, Magnetic-field-induced shape recovery by reverse phase transformation., Nature. 439 (2006) 957–60. https://doi.org/10.1038/nature04493. [39] K. Ullakko, J.K. Huang, C. Kantner, R.C.O. Handley, Large magnetic-field-induced strains in Ni2MnGa single crystals Large magnetic-field-induced strains in Ni 2 MnGa single crystals, Appl. Phys. Lett. 69 (2013) 1966–1998. [40] R.Y. Umetsu, W. Ito, K. Ito, K. Koyama, A. Fujita, K. Oikawa, T. Kanomata, R. Kainuma, K. Ishida, Anomaly in entropy change between parent and martensite phases in the Ni50Mn34In16 Heusler alloy, Scr. Mater. 60 (2009) 25–28. https://doi.org/10.1016/j.scriptamat.2008.08.022. [41] J.A. Monroe, I. Karaman, B. Basaran, W. Ito, R.Y. Umetsu, R. Kainuma, K. Koyama, Y.I. Chumlyakov, Direct measurement of large reversible magnetic-field-induced strain in Ni-Co-Mn-In metamagnetic shape memory alloys, Acta Mater. 60 (2012) 6883–6891. https://doi.org/10.1016/j.actamat.2012.07.040. [42] D.Y. Cong, S. Roth, L. Schultz, Magnetic properties and structural transformations in Ni-Co-Mn-Sn multifunctional alloys, Acta Mater. 60 (2012) 5335–5351. https://doi.org/10.1016/j.actamat.2012.06.034. [43] J. Liu, N. Scheerbaum, S. Kauffmann-Weiss, O. Gutfleisch, NiMn-based alloys and composites for magnetically controlled dampers and actuators, Adv. Eng. Mater. 14 (2012) 653–667. https://doi.org/10.1002/adem.201200038. [44] V. Sokolovskiy, A. Grünebohm, V. Buchelnikov, P. Entel, Ab initio and Monte Carlo approaches for the magnetocaloric effect in Coand in-doped Ni-Mn-Ga Heusler alloys, Entropy. 16 (2014) 4992–5019. https://doi.org/10.3390/e16094992. [45] H.B. Luo, C.M. Li, Q.M. Hu, S.E. Kulkova, B. Johansson, L. Vitos, R. Yang, Firstprinciples investigations of the five-layer modulated martensitic structure in Ni2Mn(AlxGa1-x) alloys, Acta Mater. 59 (2011) 5938–5945. https://doi.org/10.1016/j.actamat.2011.06.002. 42 Introduction [46] D. Soto, F.A. Hernández, H. Flores-Zúñiga, X. Moya, L. Mañosa, A. Planes, S. Aksoy, M. Acet, T. Krenke, Phase diagram of Fe-doped Ni-Mn-Ga ferromagnetic shapememory alloys, Phys. Rev. B - Condens. Matter Mater. Phys. 77 (2008) 1–7. https://doi.org/10.1103/PhysRevB.77.184103. [47] H.B. Xiao, C.P. Yang, R.L. Wang, V. V. Marchenkov, K. Bärner, Effect of alloying element Al substitution on Ni-Mn-Sn shape memory alloy by first-principle calculations, J. Appl. Phys. 112 (2012). https://doi.org/10.1063/1.4772618. [48] T. Kihara, X. Xu, W. Ito, R. Kainuma, M. Tokunaga, Direct measurements of inverse magnetocaloric effects in metamagnetic shape-memory alloy NiCoMnIn, Phys. Rev. B - Condens. Matter Mater. Phys. 90 (2014) 1–6. https://doi.org/10.1103/PhysRevB.90.214409. [49] T. Krenke, M. Acet, E.F. Wassermann, X. Moya, L. Mañosa, A. Planes, Martensitic transitions and the nature of ferromagnetism in the austenitic and martensitic states of Ni-Mn-Sn alloys, Phys. Rev. B - Condens. Matter Mater. Phys. 72 (2005) 1–9. https://doi.org/10.1103/PhysRevB.72.014412. [50] R.B. Helmholdt, K.H.J. Buschow, Crystallographic and magnetic structure of Ni2MnSn and NiMn2Sn, J. Less-Common Met. 128 (1987) 167–171. [51] A. Ayuela, J. Enkovaara, K. Ullakko, R.M. Nieminen, Structural properties of magnetic Heusler alloys, J. Phys. Condens. Matter. 11 (1999) 2017–2026. https://doi.org/10.1088/0953-8984/11/8/014. [52] V. Sánchez-Alarcos, V. Recarte, J.I. Pérez-Landazábal, G.J. Cuello, Correlation between atomic order and the characteristics of the structural and magnetic transformations in Ni-Mn-Ga shape memory alloys, Acta Mater. 55 (2007) 3883– 3889. https://doi.org/10.1016/j.actamat.2007.03.001. [53] A. Ghosh, K. Mandal, Effect of structural disorder on the magnetocaloric properties of Ni-Mn-Sn alloy, Appl. Phys. Lett. 104 (2014). https://doi.org/10.1063/1.4862431. [54] T. Krenke, M. Acet, E.F. Wassermann, Ferromagnetism in the austenitic and martensitic states of Ni-Mn-In alloys, (2006) 1–10. https://doi.org/10.1103/PhysRevB.73.174413. [55] A.G. Varzaneh, P. Kameli, F. Karimzadeh, H. Salamati, Effect of Heat Treatment on Martensitic Transformation of Ni47Mn40Sn13 Ferromagnetic Shape Memory Alloy Prepared by Mechanical Alloying Effect of Heat Treatment on Martensitic Transformation of Ni 47 Mn 40 Sn 13 Ferromagnetic Shape Memory Alloy Prepared b, 3 (2015). https://doi.org/10.1007/s12540-015-4537-0. [56] J. Pons, R. Santamarta, V.A. Chernenko, E. Cesari, Structure of the layered martensitic phases of Ni – Mn – Ga alloys Structure of the layered martensitic phases of Ni – Mn – Ga alloys, (2006). https://doi.org/10.1016/j.msea.2006.02.179. [57] S.L. Taylor, R.N. Shah, D.C. Dunand, Ni-Mn-Ga micro-trusses via sintering of 3Dprinted inks containing elemental powders, Acta Mater. 143 (2018) 20–29. https://doi.org/10.1016/j.actamat.2017.10.002. [58] A. You, M.A.Y. Be, I. In, Metamagnetic shape memory effect in a polycrystalline alloy, 192513 (2006) 10–13. https://doi.org/10.1063/1.2203211. [59] R.Y. Umetsu, A. Sheikh, W. Ito, B. Ouladdiaf, K.R.A. Ziebeck, T. Kanomata, R. 43 Chapter 1 Kainuma, The effect of Co substitution on the magnetic properties of the Heusler alloy Ni50Mn33Sn17, Appl. Phys. Lett. 98 (2011) 10–13. https://doi.org/10.1063/1.3548558. [60] B. Rodríguez-Crespo, D. Salazar, S. Lanceros-Méndez, V. Chernenko, Development and magnetocaloric properties of Ni(Co)-Mn-Sn printing ink, J. Alloys Compd. 917 (2022) 165521. https://doi.org/10.1016/j.jallcom.2022.165521. [61] S.S. Nambiar, B.R.N. Murthy, S. Sharma, A.A. Prasanna, A.J. Chelvane, Microstructure and Mechanical Properties of Annealed Quinary Ni-Mn-Sn-Fe-In Heusler Alloy, Eng. Sci. 17 (2022) 303–308. https://doi.org/10.30919/es8d632. [62] S. Esakki Muthu, N. V. Rama Rao, M. Manivel Raja, D.M. Raj Kumar, D. Mohan Radheep, S. Arumugam, Influence of Ni/Mn concentration on the structural, magnetic and magnetocaloric properties in Ni50-xMn37+xSn13 Heusler alloys, J. Phys. D. Appl. Phys. 43 (2010). https://doi.org/10.1088/0022-3727/43/42/425002. [63] V. Zhukova, M. Ipatov, A. Granovsky, A. Zhukov, Magnetic properties of Ni-MnIn-Co Heusler-type glass-coated microwires, J. Appl. Phys. 115 (2014) 10–13. https://doi.org/10.1063/1.4868919. [64] W. Maziarz, P. Czaja, M.J. Szczerba, L. Lityńska-Dobrzyńska, T. Czeppe, J. Dutkiewicz, Influence of Ni/Mn concentration ratio on microstructure and martensitic transformation in melt spun Ni-Mn-Sn Heusler alloy ribbons, J. Alloys Compd. 615 (2015) S173–S177. https://doi.org/10.1016/j.jallcom.2013.12.164. [65] H. Zheng, W. Wang, S. Xue, Q. Zhai, J. Frenzel, Z. Luo, Composition-dependent crystal structure and martensitic transformation in Heusler Ni-Mn-Sn alloys, Acta Mater. 61 (2013) 4648–4656. https://doi.org/10.1016/j.actamat.2013.04.035. [66] A.S. Kalbfleisch, G. Matthews, P.J. Jacques, On the influence of the cooling rate on the martensitic transformation of Ni-Mn-Sn Heusler alloys, Scr. Mater. 114 (2016) 121–124. https://doi.org/10.1016/j.scriptamat.2015.12.005. [67] T. Krenke, M. Acet, X. Moya, A. Planes, Effect of Co and Fe on the inverse magnetocaloric properties of Ni-Mn-Sn , 033903 (2007). https://doi.org/10.1063/1.2761853. [68] D.Y. Cong, S. Roth, L. Schultz, Magnetic properties and structural transformations in Ni – Co – Mn – Sn multifunctional alloys, ACTA Mater. 60 (2012) 5335–5351. https://doi.org/10.1016/j.actamat.2012.06.034. [69] D.H. Wang, C.L. Zhang, H.C. Xuan, Z.D. Han, J.R. Zhang, S.L. Tang, B.X. Gu, Y.W. Du, The study of low-field positive and negative magnetic entropy changes in, 013909 (2007). https://doi.org/10.1063/1.2752140. [70] R. Das, S. Sarma, A. Perumal, A. Srinivasan, Effect of Co and Cu substitution on the magnetic entropy change in Ni 46 Mn 43 Sn 11 alloy , 901 (2011). https://doi.org/10.1063/1.3540327. [71] F. Chen, Y.X. Tong, Y.J. Huang, B. Tian, L. Li, Y.F. Zheng, Intermetallics Suppression of g phase in Ni 38 Co 12 Mn 41 Sn 9 alloy by melt spinning and its effect on martensitic transformation and magnetic properties, 36 (2013) 81–85. https://doi.org/10.1016/j.intermet.2013.01.004. [72] H. Zheng, D. Wu, S. Xue, J. Frenzel, G. Eggeler, Martensitic transformation in rapidly 44 Introduction solidified Heusler, Acta Mater. 59 (2011) 5692–5699. https://doi.org/10.1016/j.actamat.2011.05.044. [73] A. Kitanovski, Energy Applications of Magnetocaloric Materials, Adv. Energy Mater. 10 (2020). https://doi.org/10.1002/aenm.201903741. [74] R. Gauß, G. Homm, O. Gutfleisch, The Resource Basis of Magnetic Refrigeration, 21 (2016). https://doi.org/10.1111/jiec.12488. [75] S. Zare, Free piston Stirling engines : A review, (2019). https://doi.org/10.1002/er.4533. [76] I.A. Radulov, D.Y. Karpenkov, M. Specht, T. Braun, A.Y. Karpenkov, K.P. Skokov, O. Gutfleisch, T.U. Darmstadt, A. Characterization, Heat Exchangers From MetalBonded La ( Fe , Mn , Si ) 13 H x Powder, 53 (2017) 1–7. [77] A. Haleem, M. Javaid, Additive Manufacturing Applications in Industry 4.0: A Review, J. Ind. Integr. Manag. 04 (2019) 1930001. https://doi.org/10.1142/s2424862219300011. [78] L. Chen, Y. He, Y. Yang, S. Niu, H. Ren, The research status and development trend of additive manufacturing technology, Int. J. Adv. Manuf. Technol. 89 (2017) 3651– 3660. https://doi.org/10.1007/s00170-016-9335-4. [79] R.D. Sochol, E. Sweet, C.C. Glick, S.Y. Wu, C. Yang, M. Restaino, L. Lin, 3D printed microfluidics and microelectronics, Microelectron. Eng. 189 (2018) 52–68. https://doi.org/10.1016/j.mee.2017.12.010. [80] R.C.Y. Auyeung, H. Kim, S. Mathews, N. Charipar, A. Piqué, Laser additive manufacturing of embedded electronics, Elsevier Ltd, 2017. https://doi.org/10.1016/B978-0-08-100433-3.00012-9. [81] E. Sachyani Keneth, A. Kamyshny, M. Totaro, L. Beccai, S. Magdassi, 3D Printing Materials for Soft Robotics, Adv. Mater. 33 (2021) 1–17. https://doi.org/10.1002/adma.202003387. [82] S.Y. Wu, C. Yang, W. Hsu, L. Lin, 3D-printed microelectronics for integrated circuitry and passive wireless sensors, Microsystems Nanoeng. 1 (2015) 1–9. https://doi.org/10.1038/micronano.2015.13. [83] J. Hoerber, J. Glasschroeder, M. Pfeffer, J. Schilp, M. Zaeh, J. Franke, Approaches for additive manufacturing of 3D electronic applications, Procedia CIRP. 17 (2014) 806–811. https://doi.org/10.1016/j.procir.2014.01.090. [84] S.F.S. Shirazi, S. Gharehkhani, M. Mehrali, H. Yarmand, H.S.C. Metselaar, N. Adib Kadri, N.A.A. Osman, A review on powder-based additive manufacturing for tissue engineering: Selective laser sintering and inkjet 3D printing, Sci. Technol. Adv. Mater. 16 (2015). https://doi.org/10.1088/1468-6996/16/3/033502. [85] Y. Wu, Y. Lu, M. Zhao, S. Bosiakov, L. Li, A Critical Review of Additive Manufacturing Techniques and Associated Biomaterials Used in Bone Tissue Engineering, Polymers (Basel). 14 (2022) 1–24. https://doi.org/10.3390/polym14102117. [86] X. Wang, S. Xu, S. Zhou, W. Xu, M. Leary, P. Choong, M. Qian, M. Brandt, Y.M. Xie, Topological design and additive manufacturing of porous metals for bone scaffolds and orthopaedic implants: A review, Biomaterials. 83 (2016) 127–141. 45 Chapter 1 https://doi.org/10.1016/j.biomaterials.2016.01.012. [87] D. Barba, E. Alabort, R.C. Reed, Synthetic bone: Design by additive manufacturing, Acta Biomater. 97 (2019) 637–656. https://doi.org/10.1016/j.actbio.2019.07.049. [88] Razvan Udroiu, Powder bed additive manufacturing systems and its applications, Acad. J. Manuf. Eng. 10 (2012) 122–129. [89] J.P. Kruth, L. Froyen, J. Van Vaerenbergh, P. Mercelis, M. Rombouts, B. Lauwers, Selective laser melting of iron-based powder, J. Mater. Process. Technol. 149 (2004) 616–622. https://doi.org/10.1016/j.jmatprotec.2003.11.051. [90] A. Simchi, F. Petzoldt, H. Pohl, On the development of direct metal laser sintering for rapid tooling, J. Mater. Process. Technol. 141 (2003) 319–328. https://doi.org/10.1016/S0924-0136(03)00283-8. [91] B. Dutta, S. Palaniswamy, J. Choi, L.J. Song, J. Mazumder, Additive manufacturing by direct metal deposition, Adv. Mater. Process. 169 (2011) 33–36. [92] E.C. Santos, M. Shiomi, K. Osakada, T. Laoui, Rapid manufacturing of metal components by laser forming, Int. J. Mach. Tools Manuf. 46 (2006) 1459–1468. https://doi.org/10.1016/j.ijmachtools.2005.09.005. [93] M. Rincón-iglesias, A. Delgado, N. Peřinka, E. Lizundia, S. Lanceros-méndez, Waterbased 2D printing of magnetically active cellulose derivative nanocomposites, Carbohydr. Polym. 233 (2020) 115855. https://doi.org/10.1016/j.carbpol.2020.115855. [94] L. Dai, D. Xu, Polyethylene surface enhancement by corona and chemical, 60 (2019) 1005–1010. https://doi.org/10.1016/j.tetlet.2019.03.013. [95] K.N. Al-milaji, S. Gupta, V.K. Pecharsky, R. Barua, H. Zhao, R.L. Hadimani, Differential effect of magnetic alignment on additive manufacturing of magnetocaloric particles, (2020). https://doi.org/10.1063/1.5130028. [96] K.N. Al-Milaji, R.L. Hadimani, S. Gupta, V.K. Pecharsky, H. Zhao, Inkjet Printing of Magnetic Particles Toward Anisotropic Magnetic Properties, Sci. Rep. 9 (2019) 1–9. https://doi.org/10.1038/s41598-019-52699-0. [97] C. Xu, Q. Wu, G. L’Espérance, L.L. Lebel, D. Therriault, Environment-friendly and reusable ink for 3D printing of metallic structures, Mater. Des. 160 (2018) 262–269. https://doi.org/10.1016/j.matdes.2018.09.024. [98] B.Y. Ahn, D. Shoji, C.J. Hansen, E. Hong, D.C. Dunand, J.A. Lewis, Printed origami structures, Adv. Mater. 22 (2010) 2251–2254. https://doi.org/10.1002/adma.200904232. [99] E. Peng, X. Wei, T.S. Herng, U. Garbe, D. Yu, J. Ding, Ferrite-based soft and hard magnetic structures by extrusion free-forming, RSC Adv. 7 (2017) 27128–27138. https://doi.org/10.1039/c7ra03251j. 44 Chapter 2 Chapter 2 Experimental Methods 54 Experimental methods 2.6. INK PREPARATION Once a proper powder is obtained it is then used for the ink manufacture. The goal is to make an environmentally-friendly approach and for that the polymers selected for ink fabrication were those that satisfy environmentallyfriendliness, to be bio-based, water soluble and to be widely available costeffective raw materials. Specifically, the HPC was selected for having a high processability and good film forming capacity. Specifically, the inks consist of three elements:  Metallic powder: Commercial Fe, Al and Si powder and fabricated powder of NiMn-based Heusler alloy  Matrix: a polymer that acts as a binder ⇒ Hydroxypropyl Cellulose (HPC), Collagen and Silk.  Dissolvent: Deionized H2O 2.6.1. Collagen-based ink The source of collagen for ink preparation was a salmon skin which provides a good quality collagen with good mechanical properties. The first step is to separate the collagen from the skin. For that, the scales and muscles were removed, and skin piece was further washed with water and cut into pieces of about 2 x 2 cm. Then, for fat removal from the skin, the pieces are immersed in 10% ethanol for 48h, under stirring (the solution is replaced twice a day). In order to remove non-collagenous proteins the skins are then treated with 0.1M NaOH (1:10 w/v), during 3 x 2 h. After thorough washing with water, salmon skins are dissolved in 0.5M acetic acid (1:10 w/v) during 72h, under stirring. Then the resulting mixture is centrifuged and the supernatant, containing the acid soluble collagen is further vacuum filtered to remove non-soluble impurities. Now, salmon skin collagen is recovered by salting out and after centrifugation is suspended in 0.5 M of acetic acid, dialyzed against 0.1 M acetic 55 Chapter 2 acid. Then it is freeze-dried until further use. All the extraction procedure is conducted at 4 ºC. puta 2.6.2. Silk-based ink For silk-based ink preparation we start from Bombyx mori silkworm cocoons by a standard method, named soap degumming[2]. Each cocoon is cut in half, and the internal surface is cleaned mechanically. Then, several small pieces of 1 cm2 are cut. The silk cocoon is basically composed by two elements: (i) Silk fibroin, which is the polymer we are interested in; and (ii) sericin, that acts as a glue joining the fibres. For separating the sericin from the fibroin, the pieces of cocoons are introduced in an alkaline solution of 0.05 wt%. 𝑁𝑎2𝐶𝑂3 (see Figure 2.3) and is heated up to 95ºC (above 100ºC the chains starts to break but we want to avoid this) and left 10 minutes, repeating this Degumming process one more time. In this process the Hydrogen bridges of the polymer are broken. After this process white neat fibres are obtained and they are left drying overnight. puta The second part of this ink preparation process consists on dissolving the silk fibroine (SF) to obtain a gel. To accomplish this task the dried fibres are introduced in a solution that is prepared with [4g CaCl2 x 2H2O + 4mL H2O + 1.5 ml EtOH] per 0.5g of fibroine. The ethanol helps breaking Hydrogen bridges. The solution is heated up to 80ºC and magnetically mixed at 400 rpm. This produces a solution of 8% of fibroin in water. Then the mix is introduced in a dialysis cassette (cut-off: 30kDa) and is left for 6 hours. After the salt removal the final product is a solution of water and fibroin with 2% of concentration. This solution is metastable and may have two phases: Dissolved and Precipitated. The solution is centrifuged 5min at 7500 rpm to remove any solid residue that could be in it (Figure 2.3). The dried silk is semi-crystalline and during further processing the Hydrogen bridges are broken, giving rise to an amorphous material, which will be dried. 56 Experimental methods Figure 2.3: Schematic of the route for silk preparation. The metallic ink is prepared by mixing metallic powder with SF and formic acid that improves the viscosity in order to well-distribute the metallic particles within the ink. puta 2.6.3. Cellulose-based ink The flow chart for cellulose-based ink preparation is schematically shown in Figure 2.4. The preparation of the ink consists on dissolving the HPC in distilled water by manual stirring and leaving for 24 hours at room temperature until air bubbles are removed and the HPC is completely dissolved. Then the metallic powder is added and the mix is manually stirred during 5 minutes until homogenization. Figure 2.4: The flow chart of the experimental procedure for ink preparation and bubble removal. 57 Chapter 2 The process of manual stirring introduces air bubbles within the viscous ink leaving it for non-optimal printing. Introducing the ink in a vacuum chamber removes air bubbles after 2-3 cycles of vacuuming. The water evaporation in this process is minimal (less than 5%) so the water reduction can be neglected. After vacuuming the ink is introduced into a 10mL syringe for being stored and/or to start printing. In some cases the syringe is vacuumed during 2 minutes for removing completely remaining bubbles that may be introduced in the process of filling it. 2.7. 2D PRINTING 2.7.1. Doctor Blade The first step for ink testing was done by Doctor Blade, a technique that consists on depositing a drop of ink in a substrate and then passing a sharp blade at a certain height (in our case 0.5mm). This technique provides a homogeneous film that is left drying at room temperature overnight. Figure 2.5 shows the scheme of the working principle[3]. Figure 2.5: Generic scheme for film preparation by doctor blade technique. In our case the drying temperature was roomtemperature. puta 2.7.2. Screen-printing After doctor blade tests, the first printing tests of the prepared inks were done by 2D printing using the screen-printing method (Figure 2.6). The device used for this tests is the Wenzhou Zhengchang Machinery ZH3040H. The ink is deposited in the form of chips on the screen mesh and, using a manual 58 Experimental methods scrapper, is directed to the printing pattern under which a PET substrate is placed, where the ink is deposited. Since metallic inks are magnetic, placing a permanent magnet under the substrate can orientate the particles with their magnetic easy axis along the magnetic field. To increase the amount of magnetic material, 10 layers are printed one on top of the other. Figure 2.5: Screen-printing setup (left) and detail of the ink deposition in the substrate in the printing process, where it is oriented by the permanent magnet placed below the substrate (right). puta 2.8. 3D PRINTING 3D structures are printed using the cold-extrusion technique (also known as solvent-cast printing) using Tissue Scribble printer (3dcultures). The scheme of cold-extrusion technique is shown in Figure 2.7. The metallic ink, which now presents rather viscous paste, is loaded into a 10 mL syringe attached with a tapered nozzle with exit inner diameter of 410μm. The structures are printed in a PET (Polyethylene Terephthalate) foil substrate that is stuck into a levelled glass surface. In order to ensure proper PET substrate adhesion to the glass surface small drops of water are deposited at the corners of PET substrate so a thin layer of water enters by capillarity between the substrate and the glass surface. The printing speed in this work varies from 200 mm/min to 300 mm/min, depending on the ink viscosity. The syringe extrusion rate is of 1.15 ml/hour. These values of printing speed and extrusion rate were chosen to get a balance between the extrusion rate with nozzle translation speed (printing speed). To accelerate the drying rate, a cooling fan (DC 5.5V, 0.55W, 250mm from the printer nozzle) is used so there is a continuous airflow. This ensures 59 Chapter 2 that the printed layer is going to be deposited into a dried previous layer since otherwise the layer-to-layer adhesion is not optimal. Figure 2.6: Schematic model for extrusion 3D printing technique used. 2.9. CHARACTERIZATION METHODS 2.9.1. Vibrating Sample Magnetometer (VSM) The magnetic and transformation characterization were performed by using the vibrating sample magnetometer (VSM) from MicroSense Inc. (the magnetic field and temperature ranges are 0-2 T and 100-500 K, respectively). With this characterization technique the temperature and magnetic field induced phase transformations can be revealed by performing magnetization versus temperature and magnetization versus magnetic field measurements. The working scheme of the VSM is shown in Figure 2.8. The sample is placed within a uniform magnetic field (up to 2 T) and the sample holder vibrates in the Z-axis at a given frequency and the coils placed in the ends of the electromagnets measure the signal change by a magnetic induction principle. The temperature is controlled by a continuous air flux that passes through a resistor that heats it by applying a certain voltage. For the measurements below room temperature the copper tube where the air goes to the furnace is submerged into a liquid nitrogen so the air is cooled down. 60 Experimental methods Figure 2.7: Scheme of the working principle of the VSM. 2.9.2. Superconducting Quantum Interference Device (SQUID) For high magnetic field measurements (up to 7 T) a SQUID-VSM magnetometer from Quantum Design Inc. was used. This device allows to measure smaller samples than the VSM since its sensitivity is far higher than the VSM one. The measurement temperature range is from 5K to 400K. 2.9.3. Differential Scanning Calorimetry (DSC) Differential scanning calorimetry is a thermoanalytical technique where the amount of heat required to increase the temperature of a sample and a reference is measured as a function of temperature. During the measurement a reference sample with a well-defined heat capacity over the range of the temperatures to be scanned is maintained at roughly the same temperature as the measuring sample. With this technique temperature induced phase transformations can be detected and characterized. In the present work a Mettler-Toledo DSC822e calorimeter (ramp 20 K/min), under N2 atmosphere was used. 61 Chapter 2 2.9.4. Scanning Electron Microscopy (SEM) For microstructure analysis a scanning electron microscope (SEM) was used (Hitachi TM 3000). It was equipped with an energy-dispersive X-ray spectroscopy (EDS, or also known as EDX) for composition measurements. The sample to be measured is placed in the sample holder using carbon tape, to ensure electrical conductivity between them. The grains size, ribbon thickness, morphology etc. can be studied with this technique. Composition mapping was done with the EDX measurements. 2.9.5. Magnetic field-induced adiabatic temperature change A customary developed technique of direct measurements of adiabatic temperature change was used for magnetocaloric characterization of the samples. The schematic of the device is shown in Figure 2.9 [4]. MCE measurements must be performed under adiabatic conditions which mean that the sample must be thermally insulated by: (i) covering the sample with Teflon tape to ensure thermal insulation by conduction, (ii) the chamber where the sample holder is going to be mounted is kept at constant vacuum to ensure thermal insulation by convection and (iii) the shield that has the capsule where the sample is placed is made out of Copper avoiding thermal losses by infrared radiation. For making the measurement the sample is divided in two pieces and is mounted in the sample holder so the thermocouple is embedded between the two pieces of the sample. For proper heat transfer to the thermocouple a silver paste is used. Then the sample is put in contact with the heater which is integrated in the alumina-based sample holder and insulated with Teflon tape. The sample holder is then mounted in the system between two poles of a electromagnet that provides a constant magnetic field. 62 Experimental methods Figure 2.8: Sketch of the experimental setup (a). The sample holder with heater assembly (b) [4]. puta The temperature is controlled by a heater attached to the sample holder that heats the sample by thermal contact. For cooling below the room temperature a flux of air, cooled down by liquid nitrogen, is passed through an external cavity close to the cavity where the sample holder is placed. The procedure for the measurement is to heat up the sample until full austenitic phase is reached and then cooling down to the selected temperature of measurement. Then the sample is removed from the magnetic field by a hydraulic system and the value of the temperature change is measured by the thermocouple. 3.9.5. Mechanical characterization Mechanical characterization was performed to the printed structures under compression at 1 mm/min speed using Shimadzu AGS-J 500M device. 3.9.6. X-Ray Diffraction (XRD) X-ray diffraction patterns (XRDP) are obtained using a Bruker D8 Advance diffractometer (30 kV and 20 mA, λCu = 1.5418 Å) and the FullProf suite for the refinement. 63 Chapter 2 2.9.7. Magnetic field induced entropy change calculation As explained in the introduction, the magnetic entropy change can be calculated by integrating the magnetization versus temperature curves by the Maxwell thermodynamic relationship ∆𝑆𝑚(𝑇,𝐻)= 𝑆𝑚(𝑇,𝐻)−𝑆𝑚(𝑇,0)=∫(𝜕𝑀(𝑇,𝐻′) 𝜕𝑇 )𝑑𝐻′ 𝐻 0 The calculations were done using Scilab ® software and the data obtained from the SQUID isomagnetic magnetization versus temperature measurements. 70 Ribbons of Heusler-type Magnetic Shape Memory Alloys the left panel which can indicate on a much more degree of brittle failure of the HT sample compared to the as-spun counterpart. 3.2.1.2. Transformation characteristics The measurement of the temperature dependences of low-field magnetization, M(T), is a common tool to investigate the transformation behaviour, magnetic state and determine the MT and Curie temperatures in MMSMAs [4–6]. As discussed in the section 3.1 and Ref. 3, three heat treatments were performed on as-spun ribbons in order to find the optimal one[7]: (1) 723 K 30’ + q (2) 1173 K 60’ + q (3) 1173 K 60’ + q & 723K 30’ + q The q stands for “quenching” which means a very rapid cooling the sample down to 273 K by immersing it in iced water immediately after the heat treatment. This is done in order to retain the high temperature phase state. It was found that each heat treatment has a different effect on the alloy. The first one (1) improves the chemical order in the crystal structure, whereas the second one (2) improves the structural order and reduces the surface micro-strains produced during the melt-spinning technique. The third heat treatment (3) is a combination of the two previous ones resulting in being the optimal one. Figure 3.2 depicts the low-field thermomagnetization curves for both the as-spun and heat-treated ribbons samples, which were used to select the optimal heat treatment regime. When cooling from the high temperature paramagnetic austenite, the magnetization exhibits a rapid increase due to the ferromagnetic ordering of the austenitic phase at the Curie temperature. At the martensitic transformation, which also represents a magnetic transition from the ferromagnetic austenite to the antiferromagnetic martensite [8], the magnetization rapidly drops to almost zero. This strong change in the magnetic 71 Chapter 3 state characterizes a magnetostructural first-order transformation that can be easily induced by a magnetic field, resulting in several remarkable properties, particularly a significant inverse magnetocaloric effect [6,9–11]. The thermomagnetization curves for ribbons in the as-spun and heattreated conditions at 723 K show reduced MT temperatures and enhanced temperature hysteresis (Figure 3.2). These behaviours reflect the inhomogeneous state of the samples, including composition gradients, internal stresses, defects etc., caused by the fabrication techniques, such as rapid solidification during melt-spinning. Experiments demonstrate that the compositional and structural imperfections could not be eliminated by the heat treatment at 723 K for 30 min. Figure 3.2 shows considerably better MT parameters of the samples treated at 1173 K and quenched, without additional aging or aged at 723 K. Samples in these states exhibit nearly rectangular shape M(T) loops with reduced hysteresis, approximately 30 K. In this case, the heat treatments produce a much more homogeneous state in the samples, which are also relaxed from internal stresses and exhibit fewer structural defects, as evidenced by the XRD results (see chapter 4, section 4.2.1.2.). Figure 3.2: Thermomagnetization curves for Ni40Mn42.5Co8Sn9.5 ribbons with different heat treatments. 72 Ribbons of Heusler-type Magnetic Shape Memory Alloys These ribbons are suitable for powder production since the transformation after heat treatments occurs near room temperature, making this alloy suitable for (near) room-temperature cooling. The suitability of this alloy in powder form for printing is discussed in chapter 4. 3.2.2. Mn48Ni35.5Sn8Co6.5Fe2 3.2.2.1. Composition and microstructure With the intention to increase MT temperature and increase a jump of M(T) at MT we have fabricated Ni(Co)MnSn ribbon with reduced Sn and Co and increased Ni contents and added Fe atoms, whereby the nominal composition of this MMSMA became Mn48Ni35.5Sn8Co6.5Fe2 . The actual composition of this ribbon was evaluated by EDS. Table 3.2 summarize the ribbon composition, which is an average of the measurements on 10 ribbon pieces. Table 3.2: Actual compositions of the as-spun and heat treated ribbons determined by EDS analysis. Sample Mn (at %) Ni (at %) Sn (at %) Co (at %) Fe (at %) As-spun ribbons 48.2 35.2 7.5 6.5 2.8 HT ribbons 42.3 34.0 15.0 6.3 2.3 The Table 3.2 shows that content of all elements in as-spun ribbon varies within ± 1 at% from the nominal one, which corresponds to the instrumental uncertainty of the composition determination by our EDS analysis, whereas the content of Mn and Sn in HT ribbon is modified. It is known that both a Mn reduction and increasing Sn favour a shifting the MT temperatures towards low temperatures. The cross-sectional view of ribbon is documented by SEM imaging shown in Figure 3.3(left). It reveals a ribbon thickness of about 20μm and does not show a well-pronounced columnar structure. Fig. 3.3(right) shows rather globular than well-crystalized grain microstructure with the uniform size of the individual grains equal to about 1 m in diameter. 73 Chapter 3 Figure 3.3: SEM cross-sectional view image of the ribbon (left) and grain structure observed in the shiny surface of the ribbon, which was in contact with coper wheel (right). 3.2.2.2. Transformation characteristics The first measurement that gives an insight on the transformation behaviour, the Curie temperature, and MT temperature is the low-field magnetization versus temperature curve[12][13]. Figure 3.4 shows magnetization versus temperature dependences for the as-spun and heattreated ribbons at low magnetic field. Previous works have shown that performing special heat treatments may improve phase transformation and reduce hysteresis in Heusler MMSMAs[14]. However, the heat treatments that usually work well for the alloys in Ref.[16] go detrimental with the alloy studied here, as Figure 3.4 shows. In this case, the martensitic transformation is less abrupt, and the transformation front is not smooth indicating that heat treatment results in some inhomogeneities in the sample. The latter ones are responsible for a decrease of the magnetic susceptibility of alloy. All these effects are in agreement with the composition changes outlined in Table 3.2. Mn is susceptible to evaporation and since this alloy has a high manganese content, the heat treatments can have a greater impact on the alloy stoichiometry, particularly, causing a decrease of the MT temperatures, as this trend was already mentioned. 74 Ribbons of Heusler-type Magnetic Shape Memory Alloys On the other hand, the room temperature transformation characteristics and magnetic properties in the as-spun ribbon are satisfactory. Therefore, the advantage of this alloy lies in the fact that no further treatment is necessary, reducing production costs and energy consumption. Thus, this alloy is going to be studied in its as-spun ribbon form from this point of view. The values of martensitic and austenitic transformation temperatures, TM and TA respectively, are listed in Table 3.3. Figure 3.4: Thermomagnetization curves for as-spun and heat treated ribbons. Table 3.3: Curie temperature and austenitic and martensitic transformation temperatures extracted from the thermomagnetization curves for as-spun and heat treated ribbons. Sample TCA (K) TM (K) TA (K) As-spun ribbons 423 298 331 HT ribbons 408 225 265 3.2.2.3. “Magnetic field – temperature” phase diagrams of martensitic transformation The thermomagnetization curves, M(T,H), for the as-spun ribbons recorded under different magnetic fields are shown in Figure 3.5. In this figure, the magnetic field effect on MT is seen as a displacement of the MT hysteretic loop towards lower temperatures, which is a typical behaviour for MMSMAs 75 Chapter 3 and attributed to the magnetic field induced stabilization of the austenitic phase[13]. The values of martensitic temperature (TM) and austenitic temperature (TA) are extracted from M(T,H) curves using the derivative method. They are plotted as a function of magnetic field representing quasiequilibrium phase diagram of MT shown in the inset of Figure 3.5. The data in Figure 3.5 can be approximated by almost parallel to each other straight lines having negative slope of (-2.4±0.2) K/T for TA and (-1.8±0.3) K/T for TM, determined by linear fitting. Figure 3.5: Thermomagnetization curves at different applied magnetic fields. Inset shows temperature-magnetic field phase diagram of MT. magnetocaloric 3.2.2.4. Magnetic field induced entropy change Conventional and inverse magnetocaloric effects are characterized by the isothermal entropy change, ∆𝑆𝑚(𝑇,𝐻), and/or adiabatic temperature change, ∆𝑇𝑎𝑑(𝑇,𝐻), when a magnetic field is applied or removed in isothermal conditions. By means of Maxwell thermodynamic relationships the magnetic field induced entropy change can be estimated [15]. It is a common practice 76 Ribbons of Heusler-type Magnetic Shape Memory Alloys that Maxwell relationship is applied for the first-order phase transitions. From the thermomagnetization curves at different magnetic fields (Figure 3.5) the calculation of ∆𝑆𝑚(𝑇,𝐻) can be done by numerical approximation of the Maxwell relationships, as explained in chapter 2: ∆𝑆𝑚(𝑇,𝐻)=𝑆𝑚(𝑇,𝐻)−𝑆𝑚(𝑇,0)=𝜇0∫(𝜕𝑀(𝑇,𝐻′) 𝜕𝑇 )𝑑𝐻′ 𝐻 0 Figure 3.6 shows the ∆𝑆𝑚(𝑇,H) plots calculated by using cooling and heating data from Figure 3.5 for as-spun ribbon. The curves show high maximums at both the forward and reverse MT. Figure 3.7(left) in the next subsection shows field dependence of the ∆𝑆𝑚(𝑇,H) maximums. Figure 3.6: Magnetic entropy change at MT under different magnetic fields derived from M(T) curves for as-spun ribbon. magnetocaloric 3.2.2.5. Refrigerant capacity The refrigerant capacity is the amount of heat transferred between hot and cold reservoirs in a single refrigeration cycle. It can be calculated by integrating the curve of ∆𝑆𝑚(𝑇,𝐻) at FWHM: 77 Chapter 3 𝑅𝐶≈ ∫ |∆𝑆𝑀(𝑇,𝐻)|𝑑𝑇 𝑇ℎ𝑜𝑡 𝑇𝑐𝑜𝑙𝑑 Integrating the curves from Figure 3.6 yields the refrigeration capacity versus magnetic field dependences for cooling and heating cycles depicted in Figure 3.7(right). This integration procedure is referred to as a Gschneidner refrigerant capacity (RC) calculation[16]. It is worth noting that field dependencies of the maximum of ∆𝑆𝑚(𝑇,𝐻) and RC are very similar. Figure 3.7: Maximum magnetic entropy (left) and refrigerant capacity (right) as a function of the magnetic field. magnetocaloric 3.1.2.6. Adiabatic magnetocaloric effect The adiabatic temperature change, ΔTad, was measured by rapid inserting and removing the samples from a constant magnetic field of 2 T produced by the electromagnet. These measurements have been performed at constant temperatures during step-wise heating/cooling ramps ranging between 260 K and 390 K. The results are presented in Figure 3.8, which shows that the peak position is observed close to 330 K for the as-spun ribbon, reaching a maximum value of ΔTad = │1.9│ K. The ribbon exhibits a defined peak of the inverse magnetocaloric effect (negative ΔTad(T) values resulting from the application of the magnetic field) in the vicinity of the reverse MT. This transformation corresponds to the phase change from a weakly magnetic martensitic phase to a ferromagnetic austenite. 78 Ribbons of Heusler-type Magnetic Shape Memory Alloys From Figure 3.8 it can be inferred that contrary to the occurrence of the considerable peak of adiabatic temperature change at the reverse MT, no peak is observed at the forward MT during cooling ramp, only one can see a positive ΔTad signal exhibiting a step-like anomaly at the forward MT. The heating peak is highly prominent because the structural transformation is induced by applied field, where a relatively large volume fraction of transformed material is involved. On the other hand, in the cooling process, it would be necessary to induce the reverse transformation under applied magnetic field from some portion of martensite overcoming the MT hysteresis. Therefore no peak was observed under applied magnetic field when the sample was step-wise cooled across the forward MT. Note, that in order to observe a peak of ΔTad at the forward MT, one need to do measurements under field removal. That is during cooling the system has to be magnetically ordered in partial austenitic state (with applied magnetic field) and in the moment of performing the measurement the systems should go to the disordered state of martensitic phase (with no applied field), inducing the transformation from martensite to austenite. Figure 3.8: Cooling-heating dependences of adiabatic temperature change measured under 2T for as-spun ribbon. 79 Chapter 3 3.2.3. Ni43Mn39Co7Sn11 3.2.3.1. Composition analysis The third alloy studied was the compound with a nominal composition of Ni43Mn39Co7Sn11. With this alloy composition we conducted a study of the influence of the same-regime melt-spinning processing on reproducibility of the composition content of resulting ribbons. Experimentally, we have seen that it is almost impossible to obtain two or more identical ribbon compositions due to factors such as the introduction of weight errors during the preparation of precursors and the duration of alloy melting, which leads to the subsequent evaporation of some of its constituents during the melt-spinning process. The results presented in Table 3.4 indicate that some variations in the ribbons compositions occur. Table 3.4: Actual compositions for each prepared ribbon measured by EDS. Sample Ni (at %) Mn (at %) Co (at %) Sn (at %) V0 43.1 39.5 7.2 10.4 V1 43.3 39.8 6.6 10.3 V2 44.7 37.8 7.3 10.2 V3 43.3 39.2 7.4 10.1 V6 43.6 38.6 7.5 10.4 3.2.3.2. Transformation characteristics As far as these types of alloys are highly sensitive to such changes, a notable alteration in the magnetic response is observed, as evident from the thermomagnetization curves shown in Figure 3.9. Taking into account the results of the transformation behaviour shown in Figure 3.9, we selected several ribbons from Table 3.4 and performed their heat treatment. The results of the thermomagnetization measurements are shown in Figure 3.10. According to Figure 3.10, the transformation and magnetic behaviour of the heat treated ribbon V3 corresponds to the most desirable characteristics of the present work. 86 Ribbons of Heusler-type Magnetic Shape Memory Alloys Figure 3.16: SEM cross-sectional view of as-spun Ni46Mn31Co5Ga17Fe1 (left) and SEM surface image of the ribbon shiny face which was in contact with Cu wheel (right). magnetocaloric 3.3.3.2. Transformation characteristics Figures 3.17 and 3.18 show the thermomagnetization curves for both asspun and heat-treated ribbons under low and high magnetic fields. All curves in these two figures indicate M(H) anomaly typical for a second order magnetic transition from the paramagnetic to the ferromagnetic states in the martensitic phase, which means that the Curie temperature (at about 238 K) is below a martensitic transformation. Note a well-pronounced Hopkinson peak at low magnetic field. This material would not yield a substantial magnetic entropy change (see equation (3)) since the magnitude of this change depends on the character of the magnetization drop which is very smeared in Figure 3.17. Thus, this alloy is not suitable for the objectives of this thesis. 87 Chapter 3 Figure 3.17: Thermomagnetization curves for Ni46Mn31Co5Ga17Fe1 ribbons in the as-spun and heat treated states. Low magnetic field curves are shown in more detail in the inset. Figure 3.18: Thermomagnetization curves for Ni46Mn31Co5Ga17Fe1 as-spun ribbons at different magnetic field. 3.3.4. Ni50Mn18.7Cu6.25Ga25 3.3.4.1. Composition and microstructure A nominal composition of Ni50Mn18.7Cu6.25Ga25 was also checked to get the ribbon exhibiting merged MT and Curie temperatures. The ribbon with a thickness of around 20μm was fabricated at 30 m/s of wheel speed. EDS analysis revealed actual compositions for the as-spun ribbon and heat treated one which are shown in Table 3.8. Although the compositions of as-spun and 88 Ribbons of Heusler-type Magnetic Shape Memory Alloys HT ribbons are quite similar, both of them are different from the nominal composition. Table 3.8: Actual composition for as-spun and heat treated ribbons. Sample Ni (at %) Mn (at %) Cu (at %) Ga (at %) As-spun 30 m/s 51.1 15.9 6.7 26.3 HT ribbons 52.2 15.8 6.8 25.2 The cross-sectional microstructures of both ribbons are depicted in Figure 3.19 showing irregular shaped grains. The fracture surface in the left image reflects more ductile nature of the fracture than the one on the right image. Figure 3.19: Cross-sectional microstructure of as-spun (left) and heat treated ribbons (right). 3.3.4.2. Transformation characteristics The results of thermomagnetization behaviour for the as-spun and heat treated ribbons at low and high magnetic field are shown in Figure 3.20. The effect of heat treatment on the curves presented in Figure 3.20 is almost negligible. Like in the case of ribbon from subsection 3.2.2.1, the dependences reflect an ordinary ferromagnetic transition in the martensitic state with Curie temperature of about 229 K, which means that MT occurs at high temperatures in the paramagnetic state. Note Hopkinson peaks at a low magnetic field. The second-order nature of this magnetic transition and a very smeared character of the M(T) curves mean that magnetocaloric performance of material should be low, making this ribbons not suitable for the aim of the thesis. 89 Chapter 3 Figure 3.20: Thermomagnetization curves for as-spun and heat treated ribbons. 3.3.5. Ni50Mn18Cu5Ga25Fe2 3.3.5.1. Composition analysis In order to decrease MT temperature and increase TCA, the ribbon with a nominal composition of Ni50Mn18Cu5Ga25Fe2 was designed using the data for alloy from Subsection 3.2.2. The actual composition was determined as shown in Table 3.9. Table 3.9: Actual composition for as-spun ribbon. Sample Ni (at %) Mn (at %) Cu (at %) Ga (at %) Fe (at %) As-spun 30 m/s 49.0 19.3 5.9 23.7 2.3 magnetocaloric 3.3.5.2. Transformation characteristics Thermomagnetization curves at different magnetic fields are shown in Figure 3.21. The M(T) dependence when cooling from high temperature exhibits first anomaly produced by Curie temperature in the austenite (at about 323 K), then during further cooling it shows a hysteretic loop due to MT (at about 238 K) which shifts to the higher temperatures under magnetic field. These features are typical for the ferromagnetic Ni-Mn-Ga (FSMAs), exactly 90 Ribbons of Heusler-type Magnetic Shape Memory Alloys opposite to the behaviors of MT in MMSMAs. Therefore this alloy does not fit to the objectives of the present work. Figure 3.21: Thermomagnetization curves at different magnetic fields. 3.4. NiMnIn MSMA SYSTEM 3.4.1. Ni45.2Mn36.7Co5.1In13.0 3.4.1.1. Composition analysis It is well-known that the MT and magnetic characteristics of NiMnInbased Heusler-type off-stoichiometric MMSMAs are very difficult to reproduce due to extremely high sensitivity to the composition variation and to degree of atomic order. In the present work, we selected nominal composition of such type of alloy Ni45.2Mn36.7Co5.1In13.0 and prepared ribbon, which actual composition is presented in Table 3.10. Table 3.10: Actual composition for as-spun ribbon. Sample Ni (at %) Mn (at %) Co (at %) In (at %) As-spun ribbon 46.8 35.5 5.5 12.2 91 Chapter 3 3.4.1.2. Transformation characteristics The low-field thermomagnetization curve of the as-spun ribbon, presented in Figure 3.22, shows a well-pronounced MT and Curie temperature for the ferromagnetic transition. MT occurs well below room temperature with thermal hysteresis of about 50 K. The M(T) dependence of the HT ribbon in this figure shows a reduced MT hysteresis, of about 30 K, but still it is too large. Moreover, MT is shifted to low temperatures, to about 100 K. All these factors are not in line with objectives of the present work so this material is discarded for the further investigations. Figure 3.22: Thermomagnetization curves of the as-spun and HT ribbons. 92 Ribbons of Heusler-type Magnetic Shape Memory Alloys 3.5. CONCLUSIONS In this Chapter we have prepared a number of the ribbons featuring three main classes of the Heusler-type MSMAs. Their basic characterization (transformation behaviour and magnetic properties) was mostly performed by thermomagnetization measurements. Based on the results obtained, the following ribbons were selected for the preparation of MMSMA powders and their all-round studies: Mn42.5Ni40Co8Sn9.5, Mn48Ni35.5Sn8Co6.5Fe2, Ni43Mn39Co7Sn11, Ni50Mn18.7Cu6.25Ga25. 93 Chapter 3 3.6. REFERENCES [1] I. Dubenko, T. Samanta, A. Kumar Pathak, A. Kazakov, V. Prudnikov, S. Stadler, A. Granovsky, A. Zhukov, N. Ali, Magnetocaloric effect and multifunctional properties of Ni-Mn-based Heusler alloys, J. Magn. Magn. Mater. 324 (2012) 3530–3534. https://doi.org/10.1016/j.jmmm.2012.02.082. [2] D.Y. Cong, S. Roth, L. Schultz, Magnetic properties and structural transformations in Ni – Co – Mn – Sn multifunctional alloys, ACTA Mater. 60 (2012) 5335–5351. https://doi.org/10.1016/j.actamat.2012.06.034. [3] E.C. Passamani, F. Xavier, E. Favre-Nicolin, C. Larica, A.Y. Takeuchi, I.L. Castro, J.R. Proveti, Magnetic properties of NiMn-based Heusler alloys influenced by Fe atoms replacing Mn, J. Appl. Phys. 105 (2009). https://doi.org/10.1063/1.3075835. [4] C.O. Aguilar-Ortiz, J.P. Camarillo-García, J. Vergara, P. Álvarez-Alonso, D. Salazar, V.A. Chernenko, H. Flores-Zúñiga, Effect of solidification rate on martensitic transformation behavior and adiabatic magnetocaloric effect of Ni 50 Mn 35 In 15 ribbons, J. Alloys Compd. 748 (2018) 464–472. https://doi.org/10.1016/j.jallcom.2018.03.074. [5] C.O. Aguilar-Ortiz, D. Soto-Parra, P. Álvarez-Alonso, P. Lázpita, D. Salazar, P.O. Castillo-Villa, H. Flores-Zúñiga, V.A. Chernenko, Influence of Fe doping and magnetic field on martensitic transition in Ni–Mn–Sn melt-spun ribbons, Acta Mater. 107 (2016) 9–16. https://doi.org/10.1016/J.ACTAMAT.2016.01.041. [6] P. Lázpita, M. Sasmaz, E. Cesari, J.M. Barandiarán, J. Gutiérrez, V.A. Chernenko, Martensitic transformation and magnetic field induced effects in Ni42Co8Mn39Sn11 metamagnetic shape memory alloy, Acta Mater. 109 (2016) 170–176. https://doi.org/10.1016/J.ACTAMAT.2016.02.046. [7] Y. Wang, D. Salas, T.C. Duong, B. Medasani, A. Talapatra, On the fast kinetics of B2 e L2 1 ordering in Ni-Co-Mn-In metamagnetic shape memory alloys, 781 (2019). https://doi.org/10.1016/j.jallcom.2018.12.034. [8] V. Golub, V.A. L’vov, O. Salyuk, J.M. Barandiaran, V.A. Chernenko, Magnetism of nanotwinned martensite in magnetic shape memory alloys, J. Phys. Condens. Matter. 32 (2020) 313001. https://doi.org/10.1088/1361-648X/ab7f69. [9] R. Kainuma, Y. Imano, W. Ito, Y. Sutou, H. Morito, S. Okamoto, O. Kitakami, K. Oikawa, a Fujita, T. Kanomata, K. Ishida, Magnetic-field-induced shape recovery by reverse phase transformation., Nature. 439 (2006) 957–60. https://doi.org/10.1038/nature04493. [10] T. Krenke, E. Duman, M. Acet, E.F. Wassermann, X. Moya, L. Manosa, A. Planes, Inverse magnetocaloric effect in ferromagnetic Ni-Mn-Sn alloys, Nat. Mater. 4 (2005) 450–454. https://doi.org/10.1038/nmat1395. 94 Ribbons of Heusler-type Magnetic Shape Memory Alloys [11] V.A. Chernenko, V.A. L’vov, E. Cesari, J.M. Barandiaran, Fundamentals of magnetocaloric effect in magnetic shape memory alloys, 1st ed., Elsevier B.V., 2019. https://doi.org/10.1016/bs.hmm.2019.03.001. [12] C.O. Aguilar-Ortiz, D. Soto-Parra, P. Álvarez-Alonso, P. Lázpita, D. Salazar, P.O. Castillo-Villa, H. Flores-Zúñiga, V.A. Chernenko, Influence of Fe doping and magnetic field on martensitic transition in Ni-Mn-Sn melt-spun ribbons, Acta Mater. 107 (2016) 9–16. https://doi.org/10.1016/j.actamat.2016.01.041. [13] P. Lázpita, M. Sasmaz, E. Cesari, J.M. Barandiarán, J. Gutiérrez, V.A. Chernenko, Martensitic transformation and magnetic field induced effects in Ni42Co8Mn39Sn11 metamagnetic shape memory alloy, Acta Mater. 109 (2016) 170–176. https://doi.org/10.1016/j.actamat.2016.02.046. [14] B. Rodríguez-Crespo, D. Salazar, S. Lanceros-Méndez, V. Chernenko, Development and magnetocaloric properties of Ni(Co)-Mn-Sn printing ink, J. Alloys Compd. 917 (2022) 165521. https://doi.org/10.1016/j.jallcom.2022.165521. [15] V.K. Pecharsky, K.A. Gschneidner, Magnetocaloric effect from indirect measurements: Magnetization and heat capacity, J. Appl. Phys. 86 (1999) 565–575. https://doi.org/10.1063/1.370767. [16] K.A. Gschneider, Recent develompents in magnetic refrigeration, 317 (1999) 69–76. https://doi.org/10.4028/www.scientific.net/MSF.315-317.69. 1 Chapter 4 Chapter 4 Powders of Heusler–type Magnetic Shape Memory Alloys 102 Powders of Heusler-type magnetic shape memory alloys Figure 4.4: Rietveld refinement of the X-rays diffractograms for cubic austenite in the as-spun ribbon (upper left), heattreated ribbons (upper right), as-ground powder (lower left) and heat-treated powder (lower right). 4.2.1.3. Transformation characteristics Figure 4.5 shows the thermomagnetization curves for the as-ground and heat treated powders. It is evident that the heat treated ribbon grinding process for obtaining the powder significantly degrades the martensitic transformation (red curve in Figure 3.2 versus black curve in Figure 4.5). In the Chapter 3, it was demonstrated that special heat treatments can significantly improve the magnetic properties. Consequently, the same heat treatment route is followed for the powder, as shown in Figure 4.5, where the abrupt martensitic transformation is successfully achieved. In summary, the M(T) dependencies enable us to determine the optimal heat treatment regime for both ribbons (as discussed in chapter 3) and powder. 103 Chapter 4 Additionally, it has been experimentally demonstrated that by using a heat treatment at 1173 K for 60’ + quenching, it is feasible to recover the transformation characteristics for the manually ground thin ribbon in the form of powder with particle sizes well below 38 µm. This finding is highly significant since, as previously mentioned, it is challenging to replicate transformation characteristics from bulk MMSMA into its powder form. In the next chapter, this finding will be validated through a detailed comparative study of the ribbon and printed powder as the ink filler. Figure 4.5: Thermomagnetization curves for Ni40Mn42.5Co8Sn9.5 powder with different heat treatments. 4.2.2. Mn48Ni35.5Sn8Co6.5Fe2 4.2.2.1. Transformation characteristics The grinding process of this ribbon resulted in a very dramatic effect on the martensitic transformation of the powder. According to the thermomagnetization dependences, shown in Figure 4.6, both the as-received and heat-treated powders show a very smeared MT and the reduced value of 104 Powders of Heusler-type magnetic shape memory alloys M(T) change at MT, factors which impeded us to consider this powder as candidate for printing. Figure 4.6: Thermomagnetization dependence for the Mn48Ni35.5Sn8Co6.5Fe2 heat-treated powder. Inset shows the results for as-ground powder. 4.2.3. Ni43Mn39Co7Sn11 4.2.3.1. Composition and microstructure From the set of ribbons prepared with this nominal composition, two alloys were chosen. The first one, V3(A1), was selected for a systematic study of the effects of heat treatments, and the second one, V1(A2), was chosen to investigate the composition’s effect on magnetocaloric properties. Table 4.3 shows the results on actual compositions (obtained by EDS with an instrumental uncertainty of ±0.5 at.%). Table 4.3: Actual compositions for the two alloys studied. Sample Ni (at %) Mn (at %) Co (at %) Sn (at %) A1 (V3) 43.3 39.2 7.4 10.1 A2 (V1) 43.3 39.8 6.6 10.3 105 Chapter 4 4.2.3.2. Transformation characteristics The temperature dependences of low-field magnetization, M(T), as well as calorimetric curves, presented in Figure 4.7 and Figure 4.8, respectively, were used to determine characteristic transformation temperatures, hysteresis of MT and the Curie temperature of the studied powders. Both figures show wellpronounced effects of heat treatments on the transformation behaviour, where it can be observed that heat treatments produced shifts in characteristic temperatures and a reduction of the thermal hysteresis of MT. The martensitic transformations and Curie temperatures were verified by DSC measurements, where MT from the high-temperature austenitic to low-temperature martensitic phases and vice versa are accompanied by exothermal and endothermal effects, respectively (Figure 4.8). Notably, sample A2 presents about twice larger maximum saturation magnetization value than A1, meaning that the former alloy saturates at lower field than A1, possibly, due to anisotropic effects induced by the fabrication method of powders (crystal texture or shape anisotropy). At high magnetic fields, both samples show similar values of the maximum magnetization. Figure 4.7: Magnetization versus temperature dependences for the heat-treated and as-received powders at low magnetic field. 106 Powders of Heusler-type magnetic shape memory alloys Figure 4.8: DSC curves for the heat-treated and as-received powders. Martensitic and austenitic transformation temperatures and the Curie point, were extracted from both M(T) and DSC curves. They are summarized in Table 4.4. It is evident from Table 4.4 that all the transformation temperatures measured by DSC curves correlated with those obtained from the M(T) data. Table 4.4: Transformation temperatures obtained from the low magnetic field magnetization curves by the derivative method / and from the positions of extremums on DSC curves for heat-treated and no-heat-treated powders. Powder TC, K TA, K TM, K A1 No HT 388 / 395 275 / 280 248 /242 A1 HT1 393 / 396 296 / 291 270 / 268 A1 HT2 391 / 398 296 / 293 273 / 272 A2 HT2 397 / 403 298 / 295 277 / 271 107 Chapter 4 Figure 4.9: Magnetization versus temperature dependences showing the martensitic transformation evolution with the different heat treatments performed for the powder A1. 4.2.3.3. “Magnetic field – temperature” phase diagrams of martensitic transformation The magnetisation behaviour of materials at high magnetic fields is required to evaluate their magnetocaloric performance. It is also important to check the magnetization behaviour under magnetic fields achievable by the permanent magnets, such as 1.5 T [3]. Figure 4.10 shows the magnetisation versus temperature dependences for the four powders at a field of 1.5 T. As can be noticed in this figure that the magnetisation jump at MT, denoted as ΔM, is critically affected by heat treatments, being doubled for the alloys that underwent a heat treatment with respect to the non-heat-treated one. As a result, ΔM equal to 110 Am2kg-1 was recorded for H2HT2 powder, which is, to our knowledge, a record-breaking value for any MMSMAs powders yet described in the literature. It is worth noting that the size of ΔM directly relates to the magnetocaloric performance of materials. A more direct analysis of the magnetocaloric performance of the alloys can be obtained from the analysis of 108 Powders of Heusler-type magnetic shape memory alloys thermomagnetisation, M(T), curves measured under different constant magnetic fields up to 5 T, which are presented in Figure 4.11. Following to the magnetic field–induced shift of anomalies on M(T) curves (these anomalies are produced by MT), one can obtain phase diagrams for “martensitic and austenitic transformation temperatures” as a function of the magnetic field, represented in Figure 4.12 for all the samples. In all cases, there is a negative linear dependence of the MT temperatures with increasing magnetic field. The slope of this dependence for the heat-treated powders was calculated to be –(5.0±0.2) K/T, whereas TM slope for the sample A1 without heat-treatment shows a slope of –(6.3±0.1) K/T. Figure 4.10: Magnetization versus temperature dependences for the studied powders measured under 1.5 T magnetic field. 109 Chapter 4 Figure 4.11: Thermomagnetization curves at different magnetic fields for the studied powders. Figure 4.12: Phase diagrams “MT transformation temperatures versus magnetic field” for all the powders. 110 Powders of Heusler-type magnetic shape memory alloys 4.2.3.4. Magnetocaloric effect 4.2.3.4.1. Magnetic field-induced entropy change Conventional and inverse magnetocaloric effects are characterized by the isothermal entropy change, ∆𝑆𝑚(𝑇,𝐻), and/or adiabatic temperature change, ∆𝑇𝑎𝑑(𝑇,𝐻) when a magnetic field is applied or removed in the isothermal conditions. A magnetic field induced entropy change can be estimated using Maxwell thermodynamic relationships [4]. From the measured M(T) curves at different magnetic fields (Figure 4.11), one can calculate the isothermal entropy change ∆𝑆𝑚(𝑇,H) as follows: ∆𝑆𝑚(𝑇,𝐻)= 𝑆𝑚(𝑇,𝐻)−𝑆𝑚(𝑇,0)= 𝜇0∫(𝜕𝑀(𝑇,𝐻′) 𝜕𝑇 )𝑑𝐻′ 𝐻 0 (1) Figure 4.13 shows the ∆𝑆𝑚(𝑇,H) plots calculated by using cooling and heating data from Figure 4.11 for the four powders studied. Figure 4.13: Magnetic entropy change at different magnetic fields as a function of temperature derived from cooling/heating M(T) dependences shown in Figure 4.11. 111 Chapter 4 Figure 4.13 shows that the heat-treated powders exhibit values of ∆𝑆𝑚,𝑚𝑎𝑥 ≈ 35 J·kg-1·K-1 as derived from the analysis of heating M(T,H) curves at 𝜇0∆𝐻 = 7 𝑇, whereas for the non-heat-treated sample ∆𝑆𝑚,𝑚𝑎𝑥 ≈10 J·kg-1·K-1. The ∆𝑆𝑚,𝑚𝑎𝑥 values obtained from the cooling M(T) curves are lower than those obtained from the heating ones due to a more smeared character of the forward MT in comparison with the reverse MT. For example, heating curves at 𝜇0∆𝐻 = 2 T yield ∆𝑆𝑚,𝑚𝑎𝑥 ≈ 20 J·kg-1·K-1 for the heat treated powder A1; this value falls down to 12 J·kg-1·K-1 for cooling curves at the same field. Noteworthy, the values of ∆𝑆𝑚,𝑚𝑎𝑥 for the heat treated samples obtained in the present work are comparable to those of well-known magnetocaloric materials under similar applied fields, e.g., ~18.5 J·kg-1·K-1 for Gd5(Si2Ge2)[5], for LaFe11.4Si1.6 ~19.4 J·kg-1·K-1 [6], or ~25.0 J·kg-1·K-1 for Ni40Co8Mn42.5Sn9.5 [7]. There is a remarkable difference in entropy change between the heattreated and non-heat-treated powders, as expected from the less abrupt M(T) curve with a much smaller ΔM for the non-heat-treated powder in Figure 4.13(a). For A1, there is almost no difference in entropy change between HT1 and HT2, as well as between HT2 for both A1 and A2. Figure 4.14 shows a comparison of ∆𝑆𝑚(T) dependences of the four samples under field of 1.5 T. As previously mentioned, the difference between the two heat treatments for the A1 powder is negligible in the heating curves, whereas in the cooling curves, the HTA1 sample has a slightly higher ∆𝑆𝑚,𝑚𝑎𝑥 than HT2. For the non-heat-treated alloy ∆𝑆𝑚,𝑚𝑎𝑥 is about three times smaller than the rest of the samples. 118 Powders of Heusler-type magnetic shape memory alloys ordered austenite. A tiny M(T) anomaly fully discards observation of any essential MCE response at MT. Therefore, this powder is not suitable for the goal of the thesis and will not be implemented for printing. Figure 4.22: Thermomagnetization curves under different magnetic fields for the as-milled powder. The zoomed-in low field curve is shown in the inset. 4.4. CONCLUSIONS In this chapter we have prepared powder from the ribbons that were selected in the previous chapter. The basic characterization (composition, transformation behaviour and magnetocaloric effect) was mostly performed by thermomagnetization measurements. Apart from the powdered ribbons, another powder made by gas atomization was studied and characterized. Based on the results obtained, the following powders were selected for the printable magnetocaloric ink preparation: Mn42.5Ni40Co8Sn9.5 powdered ribbons and gasatomized Ni49.8Mn36.6Sn13.6 powder. 119 Chapter 4 4.5. REFERENCES [1] D. Nath, F. Singh, R. Das, X-ray diffraction analysis by Williamson-Hall, HalderWagner and size-strain plot methods of CdSe nanoparticlesa comparative study, Mater. Chem. Phys. 239 (2020) 122021. https://doi.org/10.1016/J.MATCHEMPHYS.2019.122021. [2] V. Sánchez-Alarcos, J.I. Pérez-Landazábal, V. Recarte, I. Lucia, J. Vélez, J.A. Rodríguez-Velamazán, Effect of high-temperature quenching on the magnetostructural transformations and the long-range atomic order of Ni–Mn–Sn and Ni–Mn–Sb metamagnetic shape memory alloys, Acta Mater. 61 (2013) 4676– 4682. https://doi.org/10.1016/J.ACTAMAT.2013.04.040. [3] T. Gottschall, K.P. Skokov, M. Fries, A. Taubel, I. Radulov, F. Scheibel, D. Benke, S. Riegg, O. Gutfleisch, Making a Cool Choice: The Materials Library of Magnetic Refrigeration, Adv. Energy Mater. 9 (2019). https://doi.org/10.1002/aenm.201901322. [4] V.K. Pecharsky, K.A. Gschneidner, Magnetocaloric effect from indirect measurements: Magnetization and heat capacity, J. Appl. Phys. 86 (1999) 565–575. https://doi.org/10.1063/1.370767. [5] V.K. Pecharsky, J. Gschneidner K. A., Giant Magnetocaloric Effect in Gd5Si2Ge2, Phys. Rev. Lett. 78 (1997) 4494–4497. https://doi.org/10.1103/PhysRevLett.78.4494. [6] F.X. Hu, B.G. Shen, J.R. Sun, Z.H. Cheng, G.H. Rao, X.X. Zhang, Influence of negative lattice expansion and metamagnetic transition on magnetic entropy change in the compound LaFe11.4Si1.6, Appl. Phys. Lett. 78 (2001) 3675–3677. https://doi.org/10.1063/1.1375836. [7] B. Rodríguez-Crespo, D. Salazar, S. Lanceros-Méndez, V. Chernenko, Development and magnetocaloric properties of Ni(Co)-Mn-Sn printing ink, J. Alloys Compd. 917 (2022) 165521. https://doi.org/10.1016/j.jallcom.2022.165521. [8] K.A. Gschneider, Recent develompents in magnetic refrigeration, 317 (1999) 69–76. https://doi.org/10.4028/www.scientific.net/MSF.315-317.69. [9] F. Scheibel, C. Lauhoff, P. Krooß, S. Riegg, N. Sommer, D. Koch, K. Opelt, H. Gutte, O. Volkova, S. Böhm, T. Niendorf, O. Gutfleisch, Additive manufacturing of Ni-MnSn shape memory Heusler alloy – Microstructure and magnetic properties from powder to printed parts, Materialia. 29 (2023). https://doi.org/10.1016/j.mtla.2023.101783. [10] P. Lázpita, M. Sasmaz, E. Cesari, J.M. Barandiarán, J. Gutiérrez, V.A. Chernenko, Martensitic transformation and magnetic field induced effects in Ni42Co8Mn39Sn11 metamagnetic shape memory alloy, Acta Mater. 109 (2016) 170–176. https://doi.org/10.1016/j.actamat.2016.02.046. 1 Chapter 5 Chapter 5 Ink Production and 2D – 3D printing 123 Chapter 5 Chapter 5 Design and Fabrication of Novel Metallic Printable Materials 5.1. 2D AND 3D PRINTING OF COMMERCIAL POWDERS 5.1.1. Technique validation This is the last chapter of the thesis and the one that contains its global objective: To develop a new 2D-3D printing technique and the implementation of the selected powders to print actual magnetocaloric 2D-3D structures. The first step for metallic 2D-3D printing is to elaborate a route for developing proper inks that are suitable for printing high-quality structures with good mechanical properties, uniform layer growth and high number of printable layers. For developing such route, commercial powders were used as the first step, since they are widely available in the laboratory allowing to make a systematic study of the printing limitations, optimization, etc. Then, after establishing the route for printing quality structures, the technique will be implemented to the prepared magnetocaloric powders, available in lower quantity. The main challenge for both 2D screen-printing and 3D extrusion printing techniques is to develop a proper ink so the printing result is successful. The ink is made of three elements: 1) Metallic filler 2) Matrix 3) Dissolvent The main parameter that controls the printing quality is the viscosity. For screen 2D printing the requirements of the ink are not that strict, whereas for 124 Ink Production and 2D – 3D Printing extrusion 3D printing the requirements are more sensitive and, apart from proper viscosity, the filler/matrix volumetric proportion is also crucial. 5.1.2. Searching for a binder and solvents for eco-friendly approach using metallic powder as filler Current polymer printing methods involve the use of chemical solvents and synthetic polymers. Also, high printing temperature is required (220ºC for PLA). Since our aim is to find eco-friendly approach, we need to replace those materials by environmentally friendly alternatives. In our case, the synthetic polymer is going to be replaced by three alternatives: (i) Silk, (ii) Collagen, and (iii) Cellulose derivative. All of these alternatives are found in nature so they are widely available natural source of raw materials. The chemical solvent is going to be replaced by deionized water. 5.1.2.1. Silk-based ink Once the silk is dissolved in water, the proportion of water is about 95% so the polymer content is too low for sustaining the powder to be printed. The printing tests resulted in a spread of the filament extruded losing its original shape seconds after the deposition in the substrate, making this polymer not suitable for 3D printing. 5.1.2.2. Collagen-based ink Similarly to the silk-based ink, the collagen content after it is dissolved is quite low for being able to print metallic powders, resulting in a spread of the filament in the same manner as silk inks. 5.1.2.3. Cellulose-based ink The Cellulose derivative that we are going to use is the Hydroxypropil Cellulose since it has a good processability and has been proven to have a good film forming capacity. In this case the polymer proportion, once is dissolved in water, can be tuned by increasing to high percentage. Printing tests revealed a 125 Chapter 5 consistent deposited filaments that allow to print uniform layers, making this polymer the best option for the aim of the thesis. 5.1.3. Commercial powders and ink parameters For metallic filler, various powders available in the laboratory were used. The first approach was to use a powder that is widely available and that has very similar density compared to the magnetocaloric powders (near 8 g/cm3) in order to implement the route for magnetocaloric powders. Hence, an iron powder was selected for this purpose. In addition, various tests were performed using Aluminium (Al) and Silicon (Si) powders. Although these powders have roughly 1/3 of Iron density the printing of these materials may provide a better refinement of the printing technique. Several inks were prepared for testing, first with the screen-printer and then in the extrusion printer. The metallic/cellulose weight percentage varied from 85% to 95% and the viscosity is controlled entirely by the water content of the ink, that is, for a fixed metallic percentage the water content was varied in a search for the best printing result. The weight percentage cannot be increased arbitrarily for all the powders used since the volumetric percentage of the polymer needs to be above certain value to sustain the entire powder. In the case of Iron 85% wt. the water quantity was varied from 45% to 80% in volume (25% to 40% in weight) and the best result was 48%Vol. or 28%wt (Table 5.1). For Iron 92.5% wt. the water quantity was varied from 40% to 60% in volume (15% to 30% in weight) having the best results corresponding to 51%Vol. or 22%wt. Aluminium ink was found to be optimal with water content of 23%Vol. or 40%wt., a proportion that was optimized after varying the water content from 15% to 35% in volume (or 30% to 55% in weight). For Silicon ink, the water quantity was varied from 35% to 55%Vol. (27% to 52% in weight) and the optimal proportion was found to be 47%Vol. or 37%wt (Table 5.1). 126 Ink Production and 2D – 3D Printing Table 5.1: Selected inks prepared for printing using commercial powders. The proportion of each constituent is listed both in mass fraction and in volume fraction. Ink Constituent Density (g/ml) Mass fraction (%) Volume fraction (%) Ink 1 Fe 85% wt. Fe powder (𝜙≤10 𝜇𝑚) 7.9 62 14 Deionized water 1.0 28 48 HPC 0.5 10 38 Total 1.7 100 100 Ink 2 Fe 92.5% wt. Fe powder (𝜙≤10 𝜇𝑚) 7.9 72 21 Deionized water 1.0 22 51 HPC 0.5 6 28 Total 2.4 100 100 Ink 3 Al 85% wt. Al powder (𝜙≤45 𝜇𝑚) 2.7 51 24 Deionized water 1.0 40 52 HPC 0.5 9 23 Total 1.3 100 100 Ink 4 Si 85% wt. Si powder (𝜙≤106 𝜇𝑚) 2.3 54 29 Deionized water 1.0 37 47 HPC 0.5 9 24 Total 1.3 100 100 Table 5.2: Dried ink percentages. Ink Constituent Density (g/ml) Mass fraction (%) Volume fraction (%) Ink 1 Fe 85% wt. Fe powder (𝜙≤10 𝜇𝑚) 7.9 85 27 HPC 0.5 15 73 Total 2.5 100 100 Ink 4 Fe 92.5% wt. Fe powder (𝜙≤10 𝜇𝑚) 7.9 92.5 45 HPC 0.5 7.5 55 Total 3.7 100 100 Ink 3 Al 85% wt. Al powder (𝜙≤45 𝜇𝑚) 2.7 85 51 HPC 0.5 15 49 Total 1.6 100 100 Ink 2 Si 85% wt. Si powder (𝜙≤106 𝜇𝑚) 2.3 85 55 HPC 0.5 15 45 Total 1.5 100 100 127 Chapter 5 Table 5.1 shows the selected inks from commercial powders with the proportions of each constituent in mass and volume percentage but the printed structure will contain only the polymer and the metallic powder, so it is of great interest to know the mass and volumetric proportion of polymer versus metallic powder in the dried inks, see Table 5.2. 5.1.4. Printing commercial powders 5.1.4.1. Screen-printing The first approach for ink testing was done by 2D screen-printing. In this technique 16 x 8 mm rectangle shaped chips were printed by applying 1 and 10 layers of material to increase the magnetic content and for comparison (see Figure 5.2). To modify the surface tension of the ink and improve its wettability, the surface active agent BYK-348 was added in a proportion of 3 - 5 μl/ml. Several inks were prepared for screen-printing by varying the proportions of filler material, binder and water. Three weight fractions of metallic powder, namely, 60 wt%, 75 wt% and 85 wt% were evaluated. For the ink with the 85 wt% of filler the amount of water was varied in order to get optimal viscosity maintaining the highest possible quantity of the metallic powder. It was found that the best-working ink should contain 85 wt% of powder and 15 wt% of HPC, corresponding to the powder/dissolvent proportion of 1 g/0.45 mL. This proportion resulted in an optimal ink viscosity for the screen-printing (1000–10,000 cP). Figure 5.1: Schematic representation of screen-printing setup.