ACTAUNIVERSITATIS
UPSALIENSISUPPSALA
2019
Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1782
Structural Studies of Mn-X (X=Al,Bi): Permanent Magnetic Materialswithout Rare Earth Metals
HAILIANG FANG
ISSN 1651-6214ISBN 978-91-513-0594-3urn:nbn:se:uu:diva-379177
Dissertation presented at Uppsala University to be publicly examined in Häggsalen,Ångströmlaboratoriet, Lägerhyddsvägen 1, Uppsala, Friday, 3 May 2019 at 14:15 for thedegree of Doctor of Philosophy. The examination will be conducted in English. Facultyexaminer: Associate Professor Linda Udby (University of Copenhagen, Niels Bohr Institute).
AbstractFang, H. 2019. Structural Studies of Mn-X (X=Al, Bi): Permanent Magnetic Materialswithout Rare Earth Metals. Digital Comprehensive Summaries of Uppsala Dissertations fromthe Faculty of Science and Technology 1782. 57 pp. Uppsala: Acta Universitatis Upsaliensis.ISBN 978-91-513-0594-3.
How to generate and use electricity in a more efficient way is a major challenge for humankind.In this context, permanent magnets play an important role within a very broad range of electricpower applications. The strongest magnets used today are mainly based on alloys that containrare-earth metals, which are neither economical nor sustainable. The search for new alternativealloys with satisfactory magnetic properties is the major motivation for the investigationssummarized in this thesis. Interesting candidates for alternative rare-earth free alloys wereselected with τ-MnAl as the basis. Theoretical studies suggest that such alloys may show goodmagnetic properties after chemical modifications to optimize them. Another compound withpromising magnetic properties is MnBi, included in this study.
MnAl-Z (Z= C, B, Ga as doping elements) and MnBi compounds were synthesized throughcarefully devised high-temperature methods, followed by various milling and annealing steps.The structural phase analysis of the samples was based on X-ray and neutron diffraction. Asystematic microstructural investigation was also performed for selected samples. The phasetransitions of MnAl and MnBi during heating and cooling at different rates were studied by insitu X-ray diffraction from a synchrotron source. The magnetic properties were characterizedby various methods.
By strict control of experimental parameters, the metastable τ-MnAl was found to be directlyobtainable using a "drop synthesis” process. A cooling rate of 10 K/min yielded an almost pureferromagnetic τ-MnAl phase. A microstructural characterization of similarly synthesized MnAl-C samples revealed the presence of phase segregation, a Mn-rich region and an Al-rich grainboundary phase.
A cryomilling process was employed which decreased the particle size of the MnAl-C sample.Neutron diffraction data disclosed accompanying amorphous features, related to changes in Mnand Al atom occupancies during the milling process. A flash heating procedure regenerated thestructural ordering between Mn and Al in the structure, where the initial magnetic propertieswere recovered.
The MnBi compound was synthesized by a self-flux method in order to isolate single crystals.As for τ-MnAl, in situ diffraction studies were applied for following phase transitions and themagnetic properties were studied.
Keywords: Synthesis, Magnetism, Diffraction.
Hailiang Fang, Department of Chemistry - Ångström, Inorganic Chemistry, Box 538, UppsalaUniversity, SE-751 21 Uppsala, Sweden.
© Hailiang Fang 2019
ISSN 1651-6214ISBN 978-91-513-0594-3urn:nbn:se:uu:diva-379177 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-379177)
List of Papers
This thesis based on the following papers, which are referred to in the text by their Roman numerals.
I HL. Fang, S. Kontos, J. Ångström, J. Cedervall, P. Svedlindh, K. Gunnarsson, M. Sahlberg. Directly obtained τ-phase MnAl, a high performance magnetic material for permanent magnets. Journal of Solid State Chemistry 237 (2016) 300–306
II HL. Fang, J. Cedervall, F.J.M. Casado, Z. Matej, J. Bednarcik,
J, Ångström, P. Berastegui, M, Sahlberg. Insights into for-mation and stability of τ-MnAlZx (Z = C and B). Journal of Al-loys and Compounds 692 (2017) 198-203
III HL. Fang, J. Cedervall, D. Hedlund, S. Shafeie, S. Deledda, F.
Olsson, L. von Fieandt, J. Bednarcik, P. Svedlindh, K. Gunnars-son, M. Sahlberg. Structural, microstructural and magnetic evo-lution in cryo milled carbon doped MnAl. Scientific reports, 8 (2018) 2525.
IV S Kontos, HL Fang, JH Li, E. K Delczeg-Czirjak, S Shafeie, P
Svedlindh, M Sahlberg, K Gunnarsson. Measured and calculat-ed properties B-doped τ-phase MnAl - a rare earth free perma-nent magnet. Journal of Magnetism and Magnetic Materi-als.474 (2019) 591-598.
V HL, Fang, JH Li, S Shafeie, D Hedlund, J Cedervall, F
Ekström, C Pay Gomez, J Bednarcik, P Svedlindh, K Gunnars-son, M Sahlberg. Insights into phase transitions and magnetism of MnBi crystals synthesized from self-flux. Journal of alloys and compounds 781 (2019) 308-314.
VI S Shafeie, HL Fang, D Hedlund, A Nyberg, P Svedlindh, K Gunnarsson, M Sahlberg. One step towards MnAl-based per-manent magnets - Differences in magnetic, and microstructural properties from an intermediate annealing step during synthesis. Journal of Solid State Chemistry. Accepted for publication (2019). https://doi.org/10.1016/j.jssc.2019.03.035
Reprints were made with permission from the respective publishers. Disclaimer: Part of this thesis is based on my licentiate thesis entitled MnAl-based Alloys for Magnetic Applications (Uppsala University, 2017) In addition to the papers, a worldwide patent is also granted based on the findings in this thesis: “MNAL ALLOY, PARTICLES THEREOF, AND METHOD FOR PRODUCTION”. FANG, Hailiang; SAHLBERG, Martin Häggblad; SKARMAN, Björn. Pub. No.: WO/2019/043219; International Appli-cation No.: PCT/EP2018/073595; Publication Date: 07.03.2019
My contribution to the papers
I Planned and carried out most of the experiment work including
synthesis of the samples and structural characterization. Partici-pated in analysis of the magnetic measurements. Wrote most of the manuscript and participated in all discussions.
II Planned and carried out most of the experiment work including some of the in situ measurements. Participated in analysis of the in situ data. Wrote most of the manuscript and participated in all discussions.
III Planned and carried out most of the experiment work including
synthesis of the samples and structural characterization. Per-formed some of the in situ measurements. Participated in analy-sis of the in situ data. Wrote most of the manuscript and partici-pated in all discussions.
IV Planned and carried out the experiment work including synthe-sis of the samples and structural characterization. Wrote the ex-periment part of manuscript and participated in all discussions.
V Planned and carried out most of the experiment work including
synthesis of the samples and structural characterization. Per-formed some of the in situ measurements. Participated in analy-sis of the in situ data. Wrote most of the manuscript and partici-pated in all discussions.
VI Planned and carried out parts of the experiment work including
synthesis of the samples and structural characterization. Wrote parts of the manuscript and participated in all discussions.
Contents
Abbreviations and units ................................................................................. ix
1. Introduction ............................................................................................... 11 1.1 Background ........................................................................................ 11 1.2 Permanent magnets ............................................................................ 12 1.3 Permanent magnet motors .................................................................. 14
2. Strategy and aim of the research ............................................................... 16 2.1 General aspects ................................................................................... 16 2.2 The MnAl and MnBi phases .............................................................. 17
3. Experimental methods .............................................................................. 22 3.1 Synthesis methods .............................................................................. 22
3.1.1 Drop synthesis ............................................................................ 22 3.1.2 Self-flux ...................................................................................... 23 3.1.3 Ball milling ................................................................................. 24 3.1.4 Cryo milling ................................................................................ 25 3.1.5 Heat treatment ............................................................................. 25
3.2 Characterization methods ................................................................... 25 3.2.1 Diffraction equipment ................................................................. 25 3.2.2 Bragg’s law ................................................................................. 26 3.2.3 The structure factor ..................................................................... 26 3.2.4 Synchrotron radiation ................................................................. 26 3.2.5 Neutron diffraction ..................................................................... 27 3.2.6 The Rietveld method ................................................................... 27 3.2.7 Magnetic properties characterization .......................................... 28 3.2.8 Microstructure............................................................................. 28 3.2.9 SEM ............................................................................................ 28 3.2.10 EDS ........................................................................................... 28 3.2.11 EBSD ........................................................................................ 28 3.2.12 Thermal analysis ....................................................................... 29
4. Results and discussion .............................................................................. 30 4.1 X-ray analysis of the synthesized MnAl and MnBi samples ............. 30 4.2 Phase stabilities of MnAl and MnBi .................................................. 32
4.2.1 The influence of cooling rate on the phase formation ................ 33 4.2.2 Phase stabilities and doping ........................................................ 36
4.3 Milling ................................................................................................ 37 4.4 Microstructural analysis ................................................................. 41
5. Conclusions ............................................................................................... 48
Svensk populärvetenskaplig sammanfattning ............................................... 49
Acknowledgements ....................................................................................... 51
References ..................................................................................................... 54
Abbreviations and units
XRD X-ray powder diffraction L Lorentzian function G Gaussian function η Cauchy content
VSM Vibrating sample magnetometer PPMS Physical property measurement system
Curie temperature Néel temperature Anisotropy field Anisotropy constants
Eanisotropy Magnetocrystalline anisotropy energy Intrinsic coercivity
Intrinsic magnetic field Remnant magnetization Saturation magnetization
Demagnetizing factor Magnetic permeability of free space
Induction ( ) Maximum energy product
EDS Energy-dispersive X-ray spectroscopy EBS Electron backscatter diffraction
11
1. Introduction
1.1 Background One of the most important results of the industrial revolution being a boost for global economy was the change in energy utilization from traditional sources to steam and combustion engines. However, as seen in the long run, side effects and problems like environmental pollution and global warming are now apparent [1-4].
The electrification of society has made the world economic growth possi-ble in a much more sustainable manner [5]. In some advanced economies, such as in the Nordic countries, electricity originating from renewable sources is approaching a level of almost 70% [6-9]. Still, fossil-carbon based electricity production is a significant pollution source. The carbon footprint could be reduced through switching to renewable energy [10].
Furthermore, the global energy demand could be lowered by efficiency improvement. For instance, substituting a permanent-magnet motor for a traditional induction motor, could save 5-15% of energy [11]. Since almost 45% of the global electricity is currently consumed by electric motors, a 10% increase in motor efficiency would have a tremendous impact on the world’s energy consumption and, subsequently, on the environment [12, 13].
Such a replacement also encompasses the materials of the permanent magnets themselves, i.e. new alloys should be found that are more economi-cal and sustainable. Several factors influence the search for alternatives, where some prerequisites may be mentioned:
a) Must contain an element with magnetic spins to create ferromag-netism;
b) The saturation magnetization and critical temperature of the alloyshould be large enough;
c) The components should be easily available and economical;
d) The synthesis of the compound should be feasible;
e) Possibility for the production of magnets.
In this thesis, these issues are addressed by suggesting alloys based on man-ganese. While the element metal itself does not show ferromagnetism, its
12
alloys may do so. The candidate materials studied are MnAl and MnBi, but for MnAl also doped alloys were included, the dopants being C, B and Ga.
The following sections (1.2 and 1.3) give a physical background to the magnetic properties and how permanent magnets play a role.
1.2 Permanent magnets The research and development of permanent magnetic materials during the 20th century achieved lots of progress and achievements. Several families of permanent magnets were discovered and commercially manufactured in massive quantities each year. These permanent magnets can mainly be cate-gorized into three groups: the first group is constituted by hexagonal hard ferrites like BaFe12O19 and SrFe12O19; the second group consists of rare-earth free alloys like Alnico; the third group of magnetic materials involves rare-earth elements like Sm2Co5, Sm2Co17, Nd2Fe14B [12].
The characteristics of any magnetic material is given by its response to a magnetic field , expressed as the magnetization M, in general a nonlinear relationship. For a permanent magnet, the irreversible nonlinear response process is manifested as a magnetic hysteresis loop, as illustrated in Figure 1. is the intrinsic magnetic field inside the material, defined from= − (in SI units). In this equation, H represents an applied exter-nal magnetic field, while N expresses the demagnetization factor of the ma-terial. This factor (0<
13
ter entity describes a magnet’s ability to store magnetostatic energy. Among these parameters, some are intrinsic properties of the material, like , and , while some other parameters like , and ( ) not only are determined by the intrinsic properties of the material, but are also influenced by the microstructure of the material [13, 14]. It should be noticed that the energy product for a permanent magnet used in a specific application is de-fined by the working point of the magnet, i.e. the and values defined by the load line, implying that ( ) ( ) .
Figure 1. vs. hysteresis loop of a permanent magnet.
The coercive field of is the ability of a magnetic material to withstand an external magnetic field without being demagnetized. The coercivity is essen-tially important to a permanent magnet since it makes the hysteresis loop broad. If a strong permanent magnet is magnetized to saturation by an exter-nal magnetic field and the field is then removed, magnetic field energy is stored in the material, which can be put in use in applications like motors and generators. The majority of permanent magnet research studies have focused on coercivity. There are mainly two types of coercivity mechanisms for the permanent magnet: nucleation of reverse domains and domain-wall pinning. The magnetization process of a ferromagnetic material corresponds to a change of the magnetic domain structure when exposed to an external magnetic field.
14
The maximum energy product, which for a permanent magnet with ideal shape (for = ½) of the hysteresis loop is given by ( ) = 4⁄ , is another critically important parameter for permanent magnets. It corre-sponds to the theoretical maximum magnetic energy density that can be achieved in an application. In reality, the actual energy product will be lower than suggested by that relation, being influenced by various external factors. The value of is determined by both intrinsic and extrinsic factors; the intrinsic factor depends on material properties such as , and external fac-tors include the microstructure of the magnetic material, like the degree of crystallographic grain orientation (alignment of grain easy axes) and density. A relatively high value is necessary for a material to become a candidate permanent magnet material [15].
The coercivity of a magnet is linked to the intrinsic anisotropy field of the compound, but is in practice also influenced by other factors like micro-structural defects [16]. sets the upper limit of the coercivity and is closely related to crystal symmetry. It is also connected to the relativistic spin-orbit coupling, as found for rare-earth 4f electrons.
1.3 Permanent magnet motors Permanent magnets have been broadly utilized in the area of electric motors and generators, where the permanent magnets provide a stable magnetic field in a restricted air gap. The induction motor needs an extra excitation system based on an electromagnet to provide the magnetic field through a coil around a soft magnetic core [17]. Figure 2 illustrates the principles of a six pole permanent-magnet motor and an induction motor.
Permanent magnets have two big advantages here. Firstly, the permanent magnet provides a stable magnetic field making a continuous external elec-tric power input unnecessary. External energy input is only needed to change either the rotor’s or stator’s magnetic pole to achieve a torque between the rotor and stator in order to keep the motor rotating. Because of this, without heat generation, permanent magnet powered motors or generators have a higher energy efficiency than induction motors, by up to 5-15% [11].
Secondly, since external coils are not needed for the permanent magnet rotor, the internal space is more compact (see Fig. 2) Since the total volume of a permanent magnet motor is smaller than an induction motor at the same power output [18], an efficient, compact and powerful permanent magnet motor would be the best candidate for e.g. electrical vehicle engines or intel-ligent robots [13, 19].
15
Figure 2. Illustration of volume difference between a permanent magnet motor and an
induction motor, both with six poles at the same power output.
16
2. Strategy and aim of the research
2.1 General aspects The rare-earth permanent magnets like Nd2Fe14B, Sm2Co5 and Sm2Co17 have excellent magnetic properties and are already widely used also in the renew-able “green energy” area. However, rare-earth elements and cobalt are not sustainable from a resource supply perspective. Only a small portion the extracted elements from rare-earth ores, like Sm, Pr, Nd, Dy and Tb, are suitable for making high performance permanent magnets, leaving the ma-jority of extracted light rare earth elements as a surplus [20]. Cobalt is a very expensive element, actually less abundant than many rare-earth elements. Since it is also heavily used in high energy-density Li-batteries, the price has doubled over the past few years [21].
As the basis for this study, having the aim to look for ferromagnetic alter-natives to materials containing rare-earth elements, the focus was put on the 3d elements in the periodic table for finding candidates. As a rough guide, the basic magnetism concepts like exchange energy and the Bethe-Slater curve may be used to predict the room temperature magnetic behavior.
The "exchange force" of adjacent atoms can be expressed by the ex-change energy equation:
=−2 ∑ ∙ =−2 ∑ cos
In this equation, represents the exchange interaction between neighboring atomic magnetic moments , while represents the angle between them. Figure 3 shows the Bethe-Slater curve, which is based on the "Weiss ex-change force" theory. The curve shows a relationship between the exchange energy and the ratio / of different metals; a is the is the interatomic distance and is the radius of the 3d electron shell [22].
17
Figure 3. Bethe-Slater curve illustrating cooperative magnetism for various metals.
For some 3d metals like Fe, Co and Ni, the / ratio is quite large for which exhibits a positive value, favoring a ferromagnetic state (magnetic moments oriented in parallel) on ordering. For the earlier 3d metals, antifer-romagnetic ordering due to a negative value is found, so that the spins will attain anti-parallel coupling.
Mn is a 3d element that, as a free atom, has five 3d electrons and thus a large local atomic magnetic moment. In addition, Mn is relatively low cost and widely abundant. Although, as depicted by the Bethe-Slater curve in Figure 3, metallic manganese exhibits antiferromagnetic ordering below the critical temperature (TN = 95 K), alloying Mn with Al, Bi or Ga may yield a ferromagnetic (or ferrimagnetic) state [23, 24]. The actual alloy structure is thus of major importance. This was the impetus of the present choice of ma-terials to study, as summarized in this thesis.
2.2 The MnAl and MnBi phases The τ-phase MnAl is a ferromagnetic intermetallic compound with a tetrag-onal L10-structure (CuAu structure type, space group P4/mmm), depicted in Figure 4 [25, 26]. The Mn atoms occupy mainly the (0, 0, 0) position while the Al occupies the (½, ½, ½). As seen from the phase diagram (Fig. 6), the phase generally does not take the strict 1:1 composition, but there is more Mn than Al (~10%). That means that Mn also enters the ideal Al site by sub-stitution. Although Mn has a remarkably large atomic moment per atom, the
18
alloying with non-magnetic Al has a significant dilution effect on the mag-netic moment per volume since only half of the atoms in the material can carry a magnetic moment [13, 16]. The τ-phase MnAl shows a ferromagnetic exchange between the Mn-Mn layers. With a theoretical ( ) value of 112 kJm-3, which is twice as large as the value of hard ferrites, the MnAl permanent magnetic material properties are reasonably high and could be a good candidate for a low-cost permanent magnet [23, 27].
Figure 4. The unit cell of the tetragonal L10 structure τ-phase MnAl. The arrows on the Mn atoms indicate that the magnetic spins are oriented along the tetragonal c-axis.
Figure 5 shows the Mn-Al and Mn-Bi phase diagrams. For the Mn-Al sys-tem, when the Mn content falls in the range between 50 and 60 at %, a fer-romagnetic τ-phase could be formed, however, from the phase diagram it is observed that the τ-phase is metastable and that the formation window is narrow [28, 29]. The ferromagnetic τ-phase easily decomposes into the thermodynamic two-phase mixture of non-ferromagnetic β-Mn and γ2-phase when the material is thermally activated [30-34], slowly cooled from high temperature or when kept at a moderate temperature for a long time [35]. Carbon doping has been reported to be an effective way to stabilize the τ-phase [36]. Hence, the synthesis of τ-MnAl (doped or not) presented a major challenge to be addressed. How to stabilize the metastable τ-phase and how to obtain a suitable microstructure to optimize the magnetic properties is the focus of this thesis.
Figure 5(b) depicts the Mn-Bi phase diagram, which shows that the high- temperature phase (HTP) of MnBi forms through a peritectic reaction. When the temperature is decreased below 613 K, a phase transformation occurs yielding another structure. This low-temperature phase (LTP) of MnBi be-longs to the well-known NiAs structure type (space group P63/mmm) and
19
shows ferromagnetic properties. Contrary to HTP MnBi its stoichiometry is 1:1. Due to the peritectic reaction and the phase transition it is difficult to make pure LTP MnBi using a simple straightforward synthesis method start-ing from high temperature.
20
Figure 5. (a) A selected part of the Mn-Al phase diagram [37]; (b) Mn-Bi phase diagram, as
redrawn from [38, 39].
The materials studied in this thesis are intended for utilizing as permanent magnets. The magnetic properties depend on the phase composition, crystal
21
structure and microstructure of the sample. For instance, the saturated mag-netization is proportional to the relative amount of the ferromagnetic phase in the sample. The coercivity is typically dependent on the microstructure (before it reaches the paramagnetic threshold of single domain size). These factors have been systematically investigated to gain a deeper understanding of how to control the magnetic properties.
The strict control of synthesis conditions to obtain the selected phases is of paramount importance, especially in these cases with certain experimental obstacles as sketched above. The challenge of “fighting against thermody-namics” had to be faced.
For this reason, the initial part of the study (as presented in papers I-VI) focused on the formation tendency, thermal stability and thermodynamics of the MnAl and MnBi ferromagnetic phases. Further on, experimental at-tempts to increase the thermodynamic stability of the τ-phase through doping with other elements (like C or B) were made, and microstructure refinement methods were introduced. Phase formation and transitions could be followed by in situ diffraction methods. Lastly, pressure applied under magnetic field alignment was attempted for investigating the prospects of making a perma-nent magnet from the synthesized material.
22
3. Experimental methods
3.1 Synthesis methods All the compounds in this thesis were prepared by high temperature synthe-sis methods. Two high temperature synthesis methods were used in this work, the first method is drop synthesis, and the second method is self-flux. Some of the synthesized materials were subsequently subjected to a ball milling and annealing process (relaxation).
3.1.1 Drop synthesis Some elements like Mn (Manganese), Zn (Zinc) are highly volatile at high temperatures. To minimize the evaporation, the drop synthesis method was used during the materials synthesis process (illustrated in Figure 6) [40]. During the experiment [I, II], the non-volatile elements like Al and C, B, Ga were placed inside an alumina crucible followed by heating up to 1670±20 K as measured an infrared thermometer. Subsequently, small pieces of Mn were introduced from a side tube and forced to drop down into the melt, one by one, as to react and eventually form the correct composition by prevent-ing the evaporation of manganese. The samples were then kept for a while at the chosen temperature to ensure the formation of a homogeneous alloy, after which the induction heating was turned off. The sample then attained room temperature in about 10 min. by natural cooling.
23
Figure 6. Illustration of the drop synthesis method. The volatile component is added from the
side tube by an external magnet to move pieces to fall into the induction heated melt.
3.1.2 Self-flux Self-flux is an effective method to synthesize large single crystals. Normally, the flux would consist of a metal with a fairly low melting-point, and it also works as a reactant element in the synthesis [41, 42]. Single crystals have the advantage over powder in magnetic measurements that they are more easily characterized along the easy and hard axis directions. For the synthesis of MnBi [V], a mixture of Bi and Mn in a 4:1 atomic ratio was ground into powder. The excess Bi here serves as the reactive flux. The powder mixture was placed in a Al2O3 crucible, which was sealed and put inside a stainless-steel tube under high purity argon. The temperature program of the synthesis process is shown in Figure 7. The cycling and ramping around 719 K is con-nected to the phase diagram (Fig. 5b), concerning the formation of HT-MnBi and subsequently its peritectic reaction at 628 K. The cycling was intended for promoting the growth of larger crystals while sacrificing small ones. With Bi still in liquid form, the stainless tube was then flipped upside-down and centrifuged, on which the MnBi crystals got caught in a stainless grid acting as a sieve, to be harvested for measurements. The setup of the synthe-sis is illustrated in Figure 8.
24
Figure 7. Furnace temperature program for the self-flux synthesis of MnBi.
Figure 8. Illustration of the experimental setup for the self-flux synthesis.
3.1.3 Ball milling Permanent magnets require a high coercivity to resist demagnetization. The coercivity of a permanent magnetic material is closely related to the sam-ple’s microstructure and grain size: The smaller the grain size, the higher coercivity (within the limits of the materials intrinsic properties) [43]. The highest coercivity is reached at the size corresponding to a to single magnet-ic domain [44]. The magnetic domain is a homogeneous part of the solid within which all magnetic moments of the atoms are aligned in the same
25
direction. The formation of magnetic domains is the result of a spontaneous minimization of the magneto-static energy, at the same time with the mag-netization direction in each domain randomly oriented [19]. For the drop synthesized samples, the grain sizes are usually too large for achieving sin-gle domain grains, implying a low coercivity. In order to boost the samples’ coercivity, samples may be mechanically milled to decrease the particle and grain size.
In such attempts, ingots obtained from a drop synthesis were crushed and ground by hand in a mortar [I-II]. The coarse powder was subsequently loaded into a tungsten carbide ball mill (ball: sample mass ratio of 30:1). Acetone with a purity of (99.99%) was added into the container to maintain a liquid phase. The samples were milled for 2-14 hours, and small amounts were taken out every 2 hours.
3.1.4 Cryo milling Some metals and alloys are ductile and cannot be milled at ambient tempera-ture. However, they become brittle when cooled down to low temperature. For such materials, cryogenic milling is an alternative way to create fine particles.
In this thesis, cryogenic milling experiments were carried out by a SPEX Freezer/ Mill 6770 at liquid nitrogen temperatures applied to MnAl-C made by drop synthesis [III].
3.1.5 Heat treatment While ball milling decreases the powder particle size, it also introduces strain and defects in the samples. This has a negative effect on the saturation magnetization, and the disorder is reflected in a significant peak broadening of the diffraction patterns [32]. In order to reduce this effect and possibly recover ordering, post milling heat treatments were carried out at 823 K for 30 minutes, in a 300 mbar H2 + 300 mbar Ar atmosphere.
3.2 Characterization methods
3.2.1 Diffraction equipment The X-ray powder diffraction measurements (XRD) were performed with either a Bruker D8 diffractometer using Cu Kα1 (λ = 1.540598 Å) radiation or a Bruker D8 ADVANCE diffractometer using Cu Kα radiation. The syn-chrotron radiation X-ray powder diffraction measurements were performed either on the P02.1 beamline at the Petra III synchrotron in Hamburg, Ger-
26
many or at the I711 beamline at the MAX IV Laboratory in Lund, Sweden, operating at smaller wavelengths [III].
3.2.2 Bragg’s law Almost all solid matter is crystalline, the most stable form. The atoms in a crystal arrange themselves periodically in three dimensions (quasicrystals excluded). According to the overall symmetry, crystals can divided into dif-ferent crystal classes. The various structures are classified into structure types, where the space group and atomic positions are specified.
X-rays were discovered by Wilhelm Röntgen in 1895 [45], but it was not until 1912 that the X-ray diffraction phenomenon with crystalline materials was found by Max von Laue when he used X-rays to study a diamond crys-tal [45]. In 1912–1913, William Lawrence Bragg stated the famous “Bragg equation” 2 = to hold for X-rays [46]. In fact, the Bragg rela-tion holds for the diffraction of any type of wave, i.e. also for neutrons.
3.2.3 The structure factor The intensity of the X-rays scattered by a crystal depends on the scatterers inside, the kind of atoms and their positions. The structure of a crystal may be seen as formed by a translational repetition of the unit cell, the smallest entity carrying the whole symmetry in the form of a parallelepiped. In prin-ciple, it is possible to calculate the X-ray diffraction intensity of a crystal on a relative scale by solving the diffracted intensity of a unit cell. The funda-mental concept is arrived at by adding the waves scattered from each set of lattice planes independently, called the structure factor, , in general a complex entity. If a unit cell contains N atoms, then the scattering from them, as expressed by the structure factor, is found from a summation over all N atoms on the following form:
= ∑ 2 (ℎ + + ) Here, ℎ are the Miller indices of the set of lattice planes, indicating their orientations. The structural characteristics are expressed by and xnynzn, the scattering power and positional fractional coordinates, respectively, of the atoms 1, 2..N. The diffracted intensity is proportional to | | . 3.2.4 Synchrotron radiation After the synchrotron principle was developed by the Russian scientist Vla-dimir Veksler in 1944 [47], and the first proton synchrotron source was built in 1952, the synchrotron has been widely used in materials research for e.g.
27
crystal structure determination and phase transition characterization. Com-pared with an in-house source as an X-ray tube, synchrotron radiation has several advantages. First, its intensity is considerably higher than that of an X-ray tube. Second, the wavelength of the synchrotron X-ray can be varied and selected. For instance, hard X-ray radiation is very suitable for materials research due to its high penetration power.
3.2.5 Neutron diffraction The neutron is unique in how it interacts with matter, which makes it possi-ble to distinguish properties unattainable by other probing tools. In contrast to X-rays, neutron diffraction may distinguish between elements with similar or equal number of electrons, and discern elements with low atomic number in a matrix of heavy elements. The ability to extract information about mag-netic structure and properties is another stronghold, taken advantage of in this study.
Experiments were carried out on the PUS instrument at the JEEP-II reac-tor (IFE, Kjeller, Norway) using monochromatized neutrons (λ = 1.556 Å).
3.2.6 The Rietveld method Initially published by Dr Hugo M. Rietveld in 1969, the Rietveld method has become an indispensable tool in analyzing powder diffraction data [48]. The Rietveld method is based on experimental data over a whole intensity pro-file, to be compared with a calculated pattern in a refinement process of in-strumental and structural parameters using the least-squares method, as im-plemented in various computer programs. Previous classical methods includ-ed only the summed intensities of whole peaks. The calculated peak intensity (Ihkl ) would be as follows: = | | In this equation, is the scale factor of the calculated phase, is the multiplicity of the hkl-reflection, is the combined Lorenz- and polariza-tion factors, is the structure factor, is the preferred orientation, is the absorption factor.
The peak width in the diffraction pattern is usually described by the peak half-width, the equation is given by = U + V tan +W. The pro-file shape function is influenced by the Gaussian profile shape and Lorentzi-an profile shape simultaneously. The Pseudo-Voigt function V= L η + (1-η) G uses a linear combination of Gaussian and Lorentzian functions. Only the size broadening effect was considered during the structural refinement pro-cess for the samples (not strain).
28
3.2.7 Magnetic properties characterization A 7400 Lake Shore vibrating sample magnetometer (VSM) and a Physical Property Measurement System (PPMS) were used for the magnetic charac-terization. Magnetization versus magnetic field was measured in the field range ( ) ±9 T (±90 kOe) at a constant temperature of 300 K. Magneti-zation versus temperature was measured in the VSM by using a single stage variable temperature (SSVT) insert. The magnetization was measured at a constant magnetic field of 0.1 T (1 kOe) in the temperature range 300 – 950 K.
3.2.8 Microstructure The microstructure plays an important role for the properties of a material, and it is therefore essential to gain full insight concerning this aspect. Dif-fraction provides an overall picture of the phase composition, but for very detailed microstructural information and phase-mapping, a dedicated inves-tigation is required.
3.2.9 SEM Scanning electron microscopy (SEM) is used for morphology information of the samples; the resolution is typically from nm to μm. In SEM electrons, as emitted by a field emission gun or filament, are focused and accelerated through a sequence of electromagnetic guiding lenses, then hit the surface of the sample at a very small spot. The electrons are scattered away from the surface and collected by a detector so that the morphology of the sample surface can be depicted.
3.2.10 EDS Energy-dispersive X-ray spectroscopy (EDS), as implemented in the SEM equipment, is used for elemental analysis or X-ray mapping. When the fo-cused high-energy electron beam bombards the sample surface, characteris-tic X-rays for each element are emitted. The X-rays are collected by a detec-tor resulting in an energy spectrum. The elemental distribution morphology can be reconstructed from a combination of SEM picture and X-ray energy (mapping).
3.2.11 EBSD Electron backscatter diffraction (EBSD) is a microstructural-crystallographic characterization method for crystalline phase mapping and texture analysis in the SEM instrument. The backscattered electrons from the sample surface
29
which conform to Bragg’s law diffract to form Kikuchi bands, corresponding to crystal lattice planes of crystalline phases in the sample. This information is then used for constructing phase-maps and determining the crystalline texture in the material.
3.2.12 Thermal analysis Thermal analysis implies that the studied material is undergoing thermal cycles (cooling or heating at a fixed rate) together with an inert reference. The temperature of the sample and the reference are detected by thermocou-ples. Thermal events are detected as differences in temperature between the sample and the reference (DTA, Differential thermal analysis). Any such difference is compensated in DSC (differential scanning calorimetry) so that the necessary energy input is recorded. Enthalpy changes due to exothermic or endothermic transformations can thus be measured in DSC. The infor-mation from any one of the two methods is used to determine the thermal properties of the sample.
30
4. Results and discussion
The materials studied in this thesis are intended for application as permanent magnets. The candidate materials studied are MnAl and MnBi, in the case of MnAl, both pure and C, B and Ga doped material. MnBi was investigated undoped. The following chapters present and elaborate the results achieved in detail.
All MnAl-based samples were prepared using the drop-synthesis method described above. The MnBi samples were prepared from self-flux synthesis. All samples were fully characterized as described in the following sections.
4.1 X-ray analysis of the synthesized MnAl and MnBi samples The synthesized MnAl and MnBi samples were systematically studied by X-ray powder diffraction, where the Rietveld refinement method was employed for the structural analysis. This refinement (by the Fullprof program) pro-vides a quantitative estimation for the phase content in the sample.
Figure 9 shows the XRD patterns of pure Mn0.54Al0.46 (a) and C-doped Mn0.55Al0.45C0.02 (b) samples, respectively. The very low intensity peaks and significant peak broadening of the undoped Mn0.54Al0.46 sample indicate re-sidual strain and that the material is not well crystallized. Moreover, unwant-ed phases like β-Mn and γ2 exist in the sample as well. In contrast, the crys-tallinity of Mn0.55Al0.45C0.02 in figure 9 (b) is much improved through the carbon doping, apparent from a dramatically increased signal-to- noise ratio, and very small β and γ2 phase peaks are present.
The Fullprof quantitative phase analysis reveals that the major part (~97 wt. %) of the Mn0.55Al0.45C0.02 sample consists of the ferromagnetic τ-phase, while for undoped Mn0.54Al0.46 sample, only approximately 60% of the tar-geted τ-phase is present in the sample. This distinctive difference between the carbon doped and undoped samples indicates that carbon doping has a significant stabilization effect upon the formation of the ferromagnetic τ-phase. One explanation might be that the carbon atom, on entering the inter-stitial site (½, ½, 0) in the MnAl tetragonal structure, makes the MnAl τ-phase thermodynamically more favorable. On the other hand, the possibility
31
of a kinetic effect cannot be ruled out, based on slow carbon diffusion that hampers the transition from the metastable state.
A very important discovery for the drop-synthesized MnAl and MnAlC samples is that the metastable τ-phase could be directly formed from a high temperature liquid alloy when the sample is cooled at an appropriate cooling rate, i.e. allowing optimization.
Figure 9 (c) shows the diffraction pattern of the self-flux synthesized MnBi sample. It was confirmed by single-crystal X-ray diffraction that the self-flux sample crystallized in the NiAs-type structure. The majority of the sample consisted of LTP, but still ~8% residual HTP was observed from the refinement result. This is very similar to previous reports where MnBi was synthesized by melt spinning [49, 50].
32
Figure 9. XRD powder patterns of as synthesized (a) Mn0.54Al0.46, (b) (Mn0.55Al0.45)100C2 and (c) MnBi. The calculated patterns (black) are superimposed on the observed data and the
lowest horizontal curve is their difference (blue) on the same intensity scale. The calculated peak positions of the respective phases are indicated below the patterns as vertical bars.
4.2 Phase stabilities of MnAl and MnBi A fundamental understanding of the stability of the targeted ferromagnetic phases is essential to be able to control the phase composition and micro-structure of the materials, in order to optimise the magnetic properties. In this section, the stability and phase transitions of the τ-phase MnAl and LTP MnBi are presented and discussed, as systematically studied by in situ pow-der diffraction and thermal analysis.
33
4.2.1 The influence of cooling rate on the phase formation Although it was found that the τ-phase may form directly from high temperature, the effects of the cooling rate on the phase composition was investigated in detail by X-ray diffraction different cooling rates of the sam-ple Mn0.55Al0.45C0.02. The in situ powder diffraction results for three cases (quenched, 10 K/min and 2 K/min, all starting at 900°C) are illustrated in Figure 10. On quenching, the outcome is approximately 50% each of the high-temperature phase (ε) and τ-phase. This result stands in stark contrast to that from 10 K/min cooling, yielding almost 100% τ-phase. For the lower cooling rate of 2 K/min, significant amounts of unwanted γ2 and β phases form during the end of cooling process. This small series of experiments shows the significance of controlled cooling and that 10 K/min was optimal for the formation of the τ-phase under these experimental circumstances.
34
Figure 10. Left part: Densitometric view of the in situ powder X-ray diffractograms (λ=0.207 Å) for different cooling rates a) Quenched, c) 10 K/min e) 2 K/min.
Right part: Phase composition vs. time for different cooling rates. b).quenched. d) 10 K/min(f). Formation of the β-phase during cooling at 2 K/min.
The analysis of MnBi focused on its phase transition. Figure 11 shows the DSC analysis result (using a heating rate of 5 K/min) of the as-synthesized sample. The first peak at ~545 K corresponds to the eutectic melting of small amounts of residual Bi, while the second peak at ~630 K corresponds to the LTP → HTP transition (cf. Fig. 5).
35
Figure 11. DSC results from the initial MnBi sample, using a heating rate of 5 K/min. For
details, see text.
Figure 12 shows the in situ densitometric view of the powder X-ray diffrac-tograms of MnBi during thermal cycling between 573 K and 673 K. Four different heating and cooling rates (50 K/min, 20 K/min, 10 K/min, 5 K/min) were employed during the cycling process. In addition, the reaction was investigated both without (a) and with (b) an external magnetic field applied. The external magnetic field (~0.5 T) was provided by a commercial Nd-Fe-B permanent magnet applied close to the sample.
Figure 12. Densitometric view of the powder X-ray diffractograms (λ = 0.207 Å) recorded in
situ during thermal cycling between 573 K and 673 K at different heating/cooling rates (where intensity is shown on a greyscale). (a) Without magnetic field; (b) With magnetic field
applied. From the different heating and cooling cycles, it was concluded that the phase transition between HTP and LTP is reversible without decomposition into other phases. Contrary to expectation, neither Mn-rich nor Bi-rich phas-es were detected by the in situ powder diffraction, possibly explained by the fact that the HTP↔LTP transformation occurs at a relatively low tempera-
36
ture, limiting the diffusion rate. As shown in Figure 12 (b), the application of an external magnetic field did not influence the HTP↔LTP transition pro-cess, being much too weak to have an impact [51-54].
4.2.2 Phase stabilities and doping High temperature phase stability is very important for the potential applica-bility of a magnetic material, considering that most permanent magnets em-ployed in energy conversion are operating at elevated temperatures. As shown in the previous section, carbon doping is an effective way to stabilize the τ-phase. As boron and carbon are neighbors in the periodic table with similar atomic radii, would boron also do?
As a consequence, a comparative study was performed with in situ pow-der diffraction on MnAl samples, either undoped, or doped with carbon or boron. The samples were heated to 800°C and kept at this temperature for 30 min, while the phase transition was monitored. As indicated in Figure 13, the three kinds of samples show pronounced differences regarding stability. The undoped Mn0.54Al0.46 sample is partially decomposed into β and γ2 at the end of heating process; the boron-doped sample is almost completely decom-posed; while the τ-phase is thermally stable for the carbon doped sample. This high temperature stability experiment reveals that boron doping is de-finitively not beneficial, but rather reduces the thermal stability [55]. The deviating behavior by boron might be a result from different solubility con-ditions, since the cell parameter ratio c/a remains unaffected [II, IV].
37
Figure 13. Densitometric view of the in situ powder X-ray diffractograms (λ=0.207 Å) for the
stability experiments at 800°C: a) Mn0.55Al0.45; b) Mn0.55Al0.45B0.02; c) Mn0.55Al0.45C0.02
4.3 Milling The microstructure of the as-synthesized MnAl consisted of large grains (typically 50-100 µm), yielding a very low coercivity. Milling has proven to be an effective method to decrease the grain size and increase the coercivity of the material. It was found [I, IV] that traditional ball milling did increase the coercivity remarkably, but at the cost of a lowering of the saturated mag-netization. Furthermore, the milled powder showed flaky microstructure features, which restrict magnetic field alignment and pressing [33, 56-58]. Cryomilling was seen as an alternative to overcome those problems.
Two samples of τ-phase MnAl were cryomilled for 2 and 4 hours, respec-tively, coded as 2CM and 4CM. Drop-synthesized (DS) material was com-pared with 2CM and 4CM by investigating them with XRD and neutron powder diffraction [III]. The results are shown in Figure 14.
An apparent trend for the cryomilled samples (CM) is that the diffraction peak intensities, compared with DS, decrease and the peak width increases on milling for a longer time. A changed background as well as phase content are already seen in 2CM sample. Some relatively weak peaks (at 1.73, 2.27 and 3.47 Å−1) corresponding to the τ-phase disappear in the 4CM pattern. The strongest peaks from XRD (at 2.7 ≤ Q ≤ 3.5 Å−1) are still observed, in-dicating that the 4CM sample is maintained in a crystalline state. The inten-
38
sities of the 2CM NPD data show a general decrease, while the NPD pattern for 4CM lacks pronounced features above a similar amorphous background. A change in relative occupancy of Mn and Al atoms at the (0, 0, 0) and (1/2, 1/2, 1/2) positions could explain difference in diffraction pattern between XRD and NPD data. Joint refinements of X-ray and neutron diffraction in-tensities were carried out using the Fullprof program, taking advantage of the different scattering powers encountered in XRD vs. NPD for gaining information on the individual occupancies. The refinements yielded relative Mn and Al occupancies of the DS and 2CM samples. The 4CM data were impossible to describe by Bragg peaks. Table 1 shows the refined occupan-cies of Mn and Al atoms in space group P4/mmm. The occupancy of Al at the 1d (½, ½, ½) site changes from 85% (DS) to 66% (2CM). The Mn occu-pancy at the 1a (0, 0, 0) site shows a similar change, with decrease from 94% to 75%, while the Al occupancy for 2CM at the same site increases from 6% to 25%. The lack of peaks in the NPD pattern of the 4CM sample indicates a random occupation of Mn and Al on both sites.
39
Figure 14. Refined powder diffraction data of a-b) DS, c-d) 2CM) and e-f) 4CM.
Left column: XRD data; Right column: NPD data.
40
Table 1. Refined atomic occupancies for the DS, 2CM, 4CM and all the flash-heated samples. See text for the sample coding.
Samples Atom Site occupancy (%)
1a (0, 0, 0) 1d (1/2, 1/2, 1/2) DS Mn 94.8(8) 14.5(8) Al 5.2(2) 85.4(2) 2CM Mn 75.4(6) 33.5(7) Al 24.6(4) 66.4(3) 4CM Mn Not analyzed Not analyzed Al Not analyzed Not analyzed Flash heated Mn 95 15 Al 5 85
In order to evaluate the stability as a function of temperature, the 2CM sam-ple was studied by in situ powder X-ray diffraction, with the results shown in Figure 15. The broad poorly resolved peaks of the 2CM sample (cf. Fig. 14c) disappear on the formation of ε-phase when it is heated above 900°C (920°C hold). On cooling the sample down to room temperature again, the τ-phase reappears now, however, with narrow well-defined peaks in the dif-fraction pattern (cf. Fig. 14a). Heating the cryomilled material, above the transition to the ε-phase region and back again, obviously has a favorable effect on the material – even at very high heating and cooling rates, 50 K/min).
41
Figure 15. Densitometric view of the 2CM powder X-ray diffractograms (λ = 0.207 Å) rec-
orded in-situ (where intensity is shown as a greyscale). a) 1-D image b) surface image.
4.4 Microstructural analysis Considering its importance for various properties, the microstructure of the MnAl and MnBi samples was investigated in some detail, using LOM (light optical microscopy), SEM, EDS and EBSD.
Figure 16 shows LOM images of the drop-synthesized Mn0.55Al0.45C0.02 sample, where the contrast also has been enhanced by C-DIC. See [III] for details. The contrasts most prominent in Fig. 16a indicate different phases, in line with the result of the diffraction analysis (Fig. 9b).
Figure16. LOM images of the DS sample using a) bright-field and b) C-DIC imaging. The actual phase compositions of the dark and bright regions, were estab-lished from an EBSD characterization as presented in Figure 17. As ex-pected, the sample consists of regions of a major phase (blue) and secondary phase (green). The blue and green regions could be attributed to the τ- and ε-phase, respectively. The ε-phase was not clearly established before, neither by XRD nor NPD. This difference may emanate from the ε-phase occurring in a nano-crystalline condition. Small amounts of β and γ2 phases were de-tected by EBSD, as shown by the yellow region.
42
Figure 17. Images of the DS sample of identical regions (a) from EBSD and (b), SEM.
The result from probing the chemical composition by EDS mapping of the different regions is shown in Figure 18. Considering the elemental contents, it is apparent that the sample exhibits Mn-rich (τ-phase) and Al-rich (ε-phase) regions. This means that the Mn and Al atoms migrate and precipitate differently in the matrix and grain boundary region during the cooling pro-cess [IV, VI]. The negative effects of the different regions was, however, possible to minimize using an intermediate annealing step during synthesis yielding a more homogenous material [VI].
43
Figure 18. Analysis of the DS sample using a) SEM and EDS mapping to show the b) Mn-, c)
Al-, and d) C-distribution in different regions.
For the MnBi sample, it was shown by in situ powder diffraction that in ad-dition to LTP MnBi, small amounts of HTP MnBi and Bi were present. The EDS mapping results for the single-crystal MnBi sample are shown in Figure 19. It is notable that Mn and Bi element disperse rather uniformly. However, at step corners between different layers of the crystals, a higher concentra-tion of Bi manifests the presence of residual Bi, not completely removed during the centrifugal process.
44
Figure 19. EDS X-ray mapping of centrifuged MnBi sample showing Bi and Mn distributions
for the same selected region.
4.5 Magnetic properties The magnetic properties of a permanent magnetic material are the key when assessing if the material is suitable for a specific application. In this section, the magnetic properties of MnAl and MnBi are presented.
The MnAl system: Figure 20 shows the vs. hysteresis loops of the (Mn0.55Al0.45)100C2 drop-synthesized (DS), cryomilled (CM) and flash heated (FH) samples. Detailed magnetic properties are summarized in Table 2. The hysteresis loop of the drop-synthesized sample is fairly narrow and shows a high saturation magnetization but a low coercivity. In contrast, the cry-omilled samples (Figure 20b) have more broadened hysteresis loops, indicat-ing that the coercivity of the samples has increased significantly during the milling process. As shown in Figure 20b, the 2-hour CM sample has a much wider loop than the 4-hour sample, both having a higher saturation magneti-zation and coercivity. This is well in line with the X-ray and neutron powder diffraction results: A longer milling time promotes more severe Mn and Al occupancy changes and structural disorder. As a result, the CM MnAl-C samples exhibit deteriorated magnetic properties.
The 2-hour and 4-hour cryomilled samples were flash heated (heated at 900˚C for 1-15 minutes and then quenched) in an attempt to further improve the magnetic properties. Figures 20c and 20d show the hysteresis loops of these two samples. Compared with the CM samples (Figure 20b), both FH samples exhibit much higher saturation magnetization but at the same time lower coercivity. As demonstrated in Table 2, the 10-minute FH sample shows optimum magnetic properties, while a 1-minute FH sample has a higher coercivity but much lower saturation magnetization, and a 15-minute FH a slightly lower coercivity and saturated magnetization than the 10-minute FH sample. The effect of the flash-heating process on the magnetic properties can explained by the fact that heating removes site occupancy disorder induced by the CM process. One minute is too short to remove all of the Mn and Al site occupancy disorder, while 15 minutes are a bit too
45
long. A saturation magnetization of 542.5 kA/m and a coercivity of 95 kA/m was obtained for the 2-hour CM/5-minute FH sample. The 4-hour CM/5-minute FH sample shows a saturation magnetization of 515.5 k A/m and coercivity of 127 kA/m. The magnetic characterization for the CM and FH samples reveals that the MnAl-C system can exhibit high values of both and . The MnAl-C system has a high potential for applications if further optimization of this system were achieved.
Figure 20. Magnetization versus magnetic field for a) DS; b) 2- and 4-hour CM; note change in scales c) 2 hours CM + 1 and 5- minute FH; d) 4 hours CM + 1- and 5-minute FH samples.
The small insets are blowups around the origin.
46
Table 2. Summary of the magnetic properties of the DS, CM and FH samples. The -values are expressed as µB/f.u. calculated for Mn1.1Al0.9C0.02 using the cell vol-
ume 27.96 Å3.
(kA/m)
(µB/f.u.)
(kA/m) / (N=1/3)
(kA/m)
Drop synthesized (Mn0.55Al0.45)100C2 614.8 1.84 120.4 42.1 32 Cryo milled 2 h (2CM) 104.4 0.31 37.4 37.4 302 Flash heated 1 min 248.4 0.74 81.1 37.1 183 Flash heated 5 min 542.5 1.62 133.6 34.9 95 Flash heated 15 min 527.9 1.58 132.1 35.0 103 Cryo milled 4 h (4CM) 29.8 0.09 6.8 24.0 176 Flash heated 1 min 117.6 0.35 41.9 36.4 270 Flash heated 5 min 515.8 1.54 148.4 38.3 127 Flash heated 15 min 515.2 1.54 142.8 36.7 119
Initial magnetic field alignment and pressing experiments were attempted for the CM powder. A series of pressing experiment revealed that the CM MnAl powder has much better compaction ability comparing with the as-synthesized or ball-milled powders. However, the particles consist of many grains, oriented randomly, and no improvement of the magnetic properties were observed because of unsuccessful field alignment while pressing. Better designed magnetic field pressing experiments on anisotropic MnAl powder are needed to prove whether the materials may be used in a permanent magnet application
The MnBi system: The magnetic properties of the as synthesized MnBi samples were characterized using the PPMS; hysteresis loops at different temperatures are illustrated in figure 21. The self-flux synthesized LTP MnBi has very large grains but still contains a small amount of additional HTP phase, yielding hysteresis loops with high saturation magnetization but very low coercivity. The absence of magnetic coercivity is tentatively ex-plained by nucleation of reverse domains in regions where the HTP exists, followed by domain-wall motion in the almost defect-free LTP. Evidence of this behavior can be seen from the hysteresis loop measured at 300 K; the magnetization exhibits a step-change in magnetization from posi-tive/negative saturation magnetization [V]. As demonstrated by Figure 21, the saturation magnetization increases with decreasing temperature, from 546 kA/m at 300 K to 645 kA/m at 50 K. A crushing and milling process is deemed necessary for the self-flux synthesized MnBi sample to exhibit an
47
increase in coercivity. Another way to improve the permanent magnet prop-erties of the MnBi system would be to make a phase-pure sample of LTP.
Figure 21. Representative M-H curves at different temperatures. The inset shows a magnified
part of the low-field region.
48
5. Conclusions
In this thesis, the Mn based ferromagnetic alloys MnAl and MnBi were syn-thesized by the drop-synthesis and self-flux process, respectively. The crys-tal structures, phase-transition characteristics, high-temperature stabilities and magnetic properties were systematically studied by X-ray powder dif-fraction, in situ synchrotron radiation powder diffraction, neutron powder diffraction and magnetic measurements.
For MnAl, experiments revealed that the tetragonal τ-phase could be di-rectly formed through cooling from high temperature. Doping with carbon proved beneficial for both the formation and stability of the ferromagnetic τ-phase. EDS mapping revealed that the material consists of Mn-rich and Al-rich regions attributed to the ferromagnetic τ-phase and mainly the ε-phase, respectively.
Neither Ga nor B doping has any positive effect on the stability of the τ- phase, as judged from the lack of improvement in phase purity compared to an undoped sample. However, it is shown that fast quenching could signifi-cantly promote the formation of the τ-phase.
Three different cooling rates were evaluated by in situ powder X-ray dif-fraction for carbon-doped MnAl. It was found that a moderate cooling rate of 10 K/min is optimal to form a pure τ-phase. High temperature stability ex-periments revealed that carbon doping stabilizes the τ-phase contrary to bo-ron doping.
For MnAl to attain a beneficial combination of microstructure and phase content, cryomilling was performed. The XRD characterization revealed that cryomilling affects the peak width and intensities. All the main peaks in the NPD pattern disappear after 4 hours of cryomilling, and. This is explained by assumption that the relative occupancies on the Mn and Al sites are ran-domized during the cryomilling process. In situ heating and cooling experi-ments revealed that the disordered τ-phase initially transforms into ε-phase when heated, and transforms back into the τ-phase when the sample is cooled down. By utilizing this “flash heating process”, the disordered τ-phase got fully recrystallized, and the magnetic properties were recovered, without significantly altering the microstructure.
For the Mn-Bi system, it was shown that a self-flux process enables syn-thesis of high quality ferromagnetic LTP MnBi crystals stacked along the c-axis. In situ heating/cooling experiments proved that neither the heating nor cooling rates have a significant influence on the LTP → HTP transition.
49
Svensk populärvetenskaplig sammanfattning
Att tillverka verktyg baserade på nya material är en stor del i mänsklighetens historia. Detta syns tydligt då man tittar på de olika tidsåldrarna som namn-getts efter material som människan kunde utnyttja, till exempel stenåldern, bronsåldern och järnåldern. Med järnet kom även magnetiska material till människans kännedom. Magnetiska material har haft en stor inverkan på vår historia. Ett av de första användningsområdena för magneter var som kom-passer, något som hade en oerhörd inverkan då det möjliggjorde bättre navi-gation till sjöss.
Sedan dess har magneter haft en allt större inverkan den mänskliga tek-nologiska utvecklingen. För närvarande används magneter inom hela det teknologiska spektret. Några exempel är: generatorer i vindkraftverk, elekt-riska motorer, datalagring i hårddiskar, mikrofoner och högtalare. Det är tydligt att utan effektiva magneter skulle stora delar av det moderna sam-hället stanna. Just inom energisektorn har magnetiska material en kritisk roll i omställningen till ett mer hållbart och resurseffektivt samhälle, då många av de tekniker som används för miljövänlig energi är starkt beroende av kraftfulla permanentmagneter.
De magneter som används idag kan huvudsakligen kategoriseras i två grupper. Den första gruppen är permanentmagneter baserade på sällsynta jordartsmetaller (NdFeB och SmCo); den andra gruppen är så kallade ferriter. Utav dessa två grupper så är de sällsynta jordarts-baserade magne-terna de mest högpresterande, men även väldigt dyra. Ferriter är billiga, men har inte tillräckligt bra magnetiska egenskaper för alla typer av applikationer. Utöver priset, så har även magneter baserade på sällsynta jordartsmetaller flera andra problem. Till exempel så är utvinningen av dessa metaller totalt dominerad av Kina, och brytningen av dessa metaller leder till stor miljöför-störing och radioaktiva restprodukter.
Målet med forskningen i denna avhandling är att finna nya material, som kan fylla gapet mellan dessa två grupper av magnetiska material som an-vänds idag. Ett nytt magnetiskt material, med bättre egenskaper än ferriter, kombinerat med mindre miljöpåverkan och pris än de sällsynta jordarts-baserade magneterna, skulle vara ett mycket stort framsteg i omställningen till ett hållbart samhälle.
För att kunna nå detta mål har forskningen i denna avhandling valt att fo-kusera på syntes av nya magnetiska material baserade på mangan. Mangan är en så kallad övergångsmetall, som är både väl tillgänglig, lättbruten och
50
relativt billig. Mangan har även ett starkt inneboende magnetiskt moment, vilket gör att den skulle kunna utnyttjas inom magnetiska applikationer. Pro-blemet med mangan som metall är att den är antiferromagnetisk, och som rent ämne passar mangan därför inte för dessa applikationer. Dock, genom att legera mangan med andra metaller, som aluminium eller vismut, kan man erhålla de ferromagnetiska egenskaperna som eftersträvas i en permanent-magnet.
Inom legeringssystemet Mn-Al finns en fas, τ-MnAl, som är ferromagne-tisk och har relativt goda egenskaper för att kunna fylla gapet mellan ferriter och de sällsynta jordarts-baserade magneterna. Problemet är att τ-fasen är metastabil, och som sådan svår att tillverka.
I denna avhandling har syntesen av nya manganbaserade legeringar stude-rats i detalj. Dessa studier har fokuserats på följande aspekter: syntesmetod och fasstabilitet, kristallografisk karakterisering med röntgen- och neu-trondiffraktion, mikrostrukturens morfologi samt magnetiska egenskaper. Till exempel visar resultaten att den ferromagnetiska τ-fasen är möjlig att syntetisera direkt från smälta genom den så kallade droppsyntes metoden. Utöver detta så kan en kombination av kryo-malning och snabb uppvärm-ning och kylning aktivera MnAl provet så att det erhåller goda magnetiska egenskaper.
Utöver det vetenskapliga resultaten så har även dessa studier teknologisk relevans. Forskningen har utförts i samarbete med Höganäs AB, som har stort intresse av nya magnetiska material. Resultaten från dessa studier har lett till att ett patent har erhållits med fokus på hur man kan tillverka τ-fasen på en industriell nivå. Det har även påbörjats studier på att tillverka magneter baserade på det material som tagits fram i arbetet med denna avhandling.
Även systemet mangan-vismut har en intressant fas (MnBi), som har goda magnetiska egenskaper. Denna fas är svår att tillverka då den har relativt låg termisk stabilitet och övergår till en högtemperaturfas, som har icke önsk-värda magnetiska egenskaper, vid temperaturer över 350°C. En metod för syntes av enkristaller av MnBi har utvecklats och de magnetiska egenskap-erna av dessa kristaller har undersökts.
Förhoppningen är att de studier som presenteras i denna avhandling ska bidra till utvecklingen av nya resurseffektiva sätt att generera och omvandla elektrisk energi. Även en liten förbättring av dagens teknologi kan få stor betydelse för dagens samhälle, bland annat genom att hjälpa till att motverka global uppvärmning.
51
Acknowledgements
First and foremost, I need to express my genuine appreciation to my PhD supervisor Martin Sahlberg for his great enthusiasm, patience and inspiring guidance during my past four years. Thank you very much for leading me into the atomic reciprocal diffraction space and academic world after I switched from industry. I also highly appreciate his enterprising and positive personality that I learnt a lot from.
I need to thank all my family members for all the high expectation, pa-tience, and support over the past years during this epic journey. To my par-ents and Ya’s parents, wife and son, brother, uncles, aunts, dear grandma in the heaven, friends.
Next I would like to express thanks to bulk group members Yvonne, Mar-tin, Jonas, Johan, Dennis, Pedro, Samrand, Gustav, Simon, Vitalii, Victor, Anders, Leif, Chris, Anders and former member Katalin and Jonas, special thanks to Pedro for help me a lot in experiment. All of you have given me much help and I have learnt lot of valuable things from you over the past years. To Daniel, Peter, Klas, Mikael, Sofia, Sergey, thank you for your ex-pertise in magnetism, Daniel and Peter help a lot to me about the publication and thesis.
I also need to acknowledge all the Swegrids rare earth free permanent magnet project members: Martin, Björn, Hilmar, Daniel, Peter, Gabriella, Johan, Samrand, Vitallii, Klas, Henry, Ramya, Sofia, Per-Olof, Jan. This is a fantastic group and each member has his own special expertise to make the project be so successful through broad collaboration. It is a great pleasure to do sports together with Björn, you are a very fast and strong person, I could never catch up your speed when running along the coast. To Hilmar, I enjoy your witty humor.
At the same time, I need to acknowledge Martin, Rolf Berger and Peter Svedlindh who helped me to revise the PhD thesis, went through the details one by one with patience. Also I need to express appreciation to Yawei, Yang Yu and Qingjiang who have helped me and teached me lesson during life time. To the Chair professor Ulf Jansson at our Inorganic Chemistry program, I need to thank you for the open, collaborative, freedom, respect differences, motivated academic atmosphere that you have successfully cre-ated with big wisdom in the program. Special appreciation to Jan Davidsson and Kristina Edström for great contribution to the whole department and university, providing all the resources that needed to bring all the brilliant
52
scientists of different nationalities together under the same roof to work col-laborative with each other.
Also many thanks to inorganic group members: Gunnar, Ulf, Martin, Mats, Karin, Leif, Yvonne, Rolf, Erik, Johan, Jiheng, Sarmad, Linus, Samrand, Johan Gerdin, Charlotte, Paulius, Kristina, Stefan, Ruijun, Leon, Hanna, Amirhossein, Annika, Paulius, David, Kothai, Mamoun, Zhaohui, Shuainan, Victor Manuel, Maria, Sebastian, Mikael, Kristian, Pedro, Katalin, Fang, Dennis, Samrand, Mikael, Katalin, Gustav, Markus, Yu-Kai, Anna, Aishwarya, Simon, Yiming, Barbara, Amanda, Kothai, Chris. All of you are very kind and warm persons. Also special thanks to Gunnar for the well-organized weekly Wednesday afternoon seminars.
I also need to express great thanks to the Chinese friends I meet here in Uppsala: Bo Tian, Wen Huang, Dou, Zhaohui, Ruijun, Jiji, Zhen Qiu, Junx-in, Xincang, Xiaorong, Lizhang, Le Fu, Keqiang, Shuangshuang, Rui, Hai-dong, Chunze Changgang, Lei Tian, Shihuai,, Jiaojiao, Libo, Jun Luo, Fang, Yue Ma, Hao Huang, Shenyang, Xi Chen, Xinxin, Kai Song, Chenyu, Shiyu, Qitao, Xiuquan, Mingkai, Yu Zhang, Renbing, Yu Zhang, Chengjuan, Jie Zhao, Yu Zhang, Shuainan, Bushan, Wenqing, Chengyang, Chao Xu, Hai-dong, Liyang, Ruijun, Qiuyue, Huiying, Shuyi. Besides, I need to acknowledge all the "Lunch brainstorm members" of Zhen Zhang, Jiefang, Zhibing, Zhaohui, Yu-Kai, Wei-Chao, Shengyang, Huan Wang, Xinxin, Zhenhuang, Dou, Ruijun, Mingkai, Haidong, Zhigang, we have a lot of great discussion about history, philosophy, etc. that expand the width and depth of everyday life to make it vivid.
Also I need to thank Anders, Mikael and Henrik, Håkan for their nice technical support and help to assist me in my experiments. Thanks to my roommate Yu-Chuan, Samrand and Lasse, it’s a great time to share office with you, great talks with you. Also special thanks to my former desk-mate Ronnie, a very smart person. Thanks to structure group Kristina, Torbjörn, Kersti, Daniel, Fredrik, Josh, Ivar, Matthew, Anders, Chao, Roland, Henrik, Ali, Dou, Anna, Tim, Meysam, Julia Morat, Mario, Anti, Yue, Viktor Nils-son, Antonia, Julia Maibach, Ashok, Habtom, Pavlin, Cesar, Kim, Reza, Erik, Ronnie, Chenjuan, Matthew Lacey, Håkan, Getachew, Tim, Will, Ste-ven, Erik Berg, Zhigang, Simon Colbin, Andy, Simon, Taha, Jolla, Jonas Mindemark, Victor Renman, Funeka, Siham, Jiefang, Burak, Wei, Rassmus, Girma, Harish, Olof, Shuang, Guiomar, Hohyoun, Muhammad, Solveig, Fabian, Therese, Dickson, Andreas, Girish, Pushpaka, Yutaro, Yunqi, Ann Mari, Roland, Yonas, Tesfamhret, Mikael, Chunze, Chao Zhang, Shweta, Jonas Ångström, Jonas, Krishna Akshay, Christofer, Mahsa, Le Anh, Ida, Hanna, Rickard, Robin, Isabell, Bojana, Tatiana, Jolla, Aatto, Ming-Tao, Tian Khoon, Haidong, Amber, Jonas Mindermark, Jonas Hedman, Jonas Hailemariam, Prithwiraj, Alma, Ageo, Nataliia, Andrew, Andrew, special thanks to the structural chemistry chair professor of Daniel Brandell who is an open minded professor. Thanks to theoretical chemistry member of Ro-
53
land, Ignacio, Lasse, Morgena, Marcus, Meiyuan, Erik. Thanks to the previ-ous and current and administration members for your great help, Diana, Eva, Susanne, Jennie, Camila, Terese, Nina, Anna, Tatti, Slavica and Kristoffer, Lina, Anton, also thanks to the IT service help from Peter and Patrik.
Thanks to “Jobbgympagruppen” for the nice Monday weekly sport activi-ties, Pedro, Martin, Johan Cedervall, Johan Gerdin, Andreas, Steven, Mark, David, Sarmad, Sebastian, Victor, Steven, Dennis, Will, Simon, etc.
At last, thanks a lot to Leif for his nice comments and revision of my lic-centiate and the thesis, and also thanks a lot for your work as the program guardian, I need to pay tribute to his upright personality. At the same time, great thanks to my licentiate opponent Gabriella Andersson and PhD disser-tation opponent Linda Udby.
54
References
1. F. Barbir, T.N. Veziroǧlu, H.J. Plass, Environmental damage due to fossil fuels use, International Journal of Hydrogen Energy 15(10) (1990) 739-749.
2. J.-K. He, Objectives and strategies for energy revolution in the context of tack-ling climate change, Advances in Climate Change Research 6(2) (2015) 101-107.
3. T. Sueyoshi, M. Goto, Returns to scale and damages to scale on U.S. fossil fuel power plants: Radial and non-radial approaches for DEA environmental as-sessment, Energy Economics 34(6) (2012) 2240-2259.
4. W. Krewitt, T. Heck, A. Trukenmüller, R. Friedrich, Environmental damage costs from fossil electricity generation in Germany and Europe, Energy Policy 27(3) (1999) 173-183.
5. B.K. Sovacool, How long will it take? Conceptualizing the temporal dynamics of energy transitions, Energy Research & Social Science 13 (2016) 202-215.
6. A.T. Gullberg, D. Ohlhorst, M. Schreurs, Towards a low carbon energy future – Renewable energy cooperation between Germany and Norway, Renewable En-ergy 68 (2014) 216-222.
7. A. Aslani, M. Naaranoja, K.-F.V. Wong, Strategic analysis of diffusion of re-newable energy in the Nordic countries, Renewable and Sustainable Energy Reviews 22 (2013) 497-505.
8. F. Urban, J. Nordensvärd, Low Carbon Energy Transitions in the Nordic Coun-tries: Evidence from the Environmental Kuznets Curve, 11(9) (2018) 2209.
9. R. Fischer, E. Elfgren, A. Toffolo, Energy Supply Potentials in the Northern Counties of Finland, Norway and Sweden towards Sustainable Nordic Electrici-ty and Heating Sectors: A Review, 11(4) (2018) 751.
10. C. Zou, Q. Zhao, G. Zhang, B. Xiong, Energy revolution: From a fossil energy era to a new energy era, Natural Gas Industry B 3(1) (2016) 1-11.
11. J. Kang, General purpose permanent magnet motor drive without speed and position sensor, Yaskawa Electric Corp., Japan, Yaskawa Application Note WP. AFD (2009).
12. O. Gutfleisch, M.A. Willard, E. Brück, C.H. Chen, S.G. Sankar, J.P. Liu, Mag-netic Materials and Devices for the 21st Century: Stronger, Lighter, and More Energy Efficient, Advanced Materials 23(7) (2011) 821-842.
13. R.W. McCallum, L.H. Lewis, R. Skomski, M.J. Kramer, I.E. Anderson, Practi-cal Aspects of Modern and Future Permanent Magnets, Annual Review of Ma-terials Research 44(1) (2014) 451-477.
14. F. Josef, S. Thomas, H. Sabine, H. Maciej, Recent developments in hard mag-netic bulk materials, Journal of Physics: Condensed Matter 16(5) (2004) S455.
15. R. Skomski, J. Coey, Magnetic anisotropy—How much is enough for a perma-nent magnet?, Scripta Materialia 112 (2016) 3-8.
16. J.M.D. Coey, Hard Magnetic Materials: A Perspective, IEEE Transactions on Magnetics 47(12) (2011) 4671-4681.
55
17. J.M.D. Coey, Permanent magnet applications, Journal of Magnetism and Mag-netic Materials 248(3) (2002) 441-456.
18. J. Coey, Industrial applications of permanent magnetism, Physica Scripta 1996(T66) (1996) 60.
19. J.M. Coey, Magnetism and magnetic materials, Cambridge University Press 2010.
20. S. Hoenderdaal, L. Tercero Espinoza, F. Marscheider-Weidemann, W. Graus, Can a dysprosium shortage threaten green energy technologies?, Energy 49 (2013) 344-355.
21. D. Larcher, J.M. Tarascon, Towards greener and more sustainable batteries for electrical energy storage, Nature Chemistry 7 (2014) 19.
22. S. Blundell, D. Thouless, Magnetism in condensed matter, Oxford university press New York2001.
23. J.M. Coey, New permanent magnets; manganese compounds, Journal of phys-ics. Condensed matter 26 (6) (2014) 064211.
24. J. Yang, W. Yang, Z. Shao, D. Liang, H. Zhao, Y. Xia, Y. Yang, Mn-based permanent magnets, Chinese Physics B 27(11) (2018) 117503.
25. A.J.J. Koch, P. Hokkeling, M.G. v. d. Steeg, K.J. de Vos, New Material for Permanent Magnets on a Base of Mn and Al, Journal of Applied Physics 31(5) (1960) S75-S77.
26. H. Kono, On the ferromagnetic phase in manganese-aluminum system, Journal of the Physical Society of Japan 13(12) (1958) 1444-1451.
27. J.H. Park, Y.K. Hong, S. Bae, J.J. Lee, J. Jalli, G.S. Abo, N. Neveu, S.G. Kim, C.J. Choi, J.G. Lee, Saturation magnetization and crystalline anisotropy calcula-tions for MnAl permanent magnet, Journal of Applied Physics 107(9) (2010) 09A731.
28. L. Pareti, F. Bolzoni, F. Leccabue, A.E. Ermakov, Magnetic anisotropy of MnAl and MnAlC permanent magnet materials, Journal of Applied Physics 59(11) (1986) 3824-3828.
29. K.H. Han, C.T. Lee, W.K. Choo, On the Position of Carbon Atom in the τ-Phase of Carbon-Doped Mn-Al Permanent Magnets, physica status solidi (a) 136(1) (1993) 21-28.
30. J.-G. Lee, X.-L. Wang, Z.-D. Zhang, C.-J. Choi, Effect of mechanical milling and heat treatment on the structure and magnetic properties of gas atomized Mn–Al alloy powders, Thin Solid Films 519(23) (2011) 8312-8316.
31. O. Obi, L. Burns, Y. Chen, T. Fitchorov, S. Kim, K. Hsu, D. Heiman, L.H. Lewis, V.G. Harris, Magnetic and structural properties of heat-treated high-moment mechanically alloyed MnAlC powders, Journal of Alloys and Com-pounds 582 (2014) 598-602.
32. Q. Zeng, I. Baker, J.B. Cui, Z.C. Yan, Structural and magnetic properties of nanostructured Mn–Al–C magnetic materials, Journal of Magnetism and Mag-netic Materials 308(2) (2007) 214-226.
33. H. Jian, K.P. Skokov, O. Gutfleisch, Microstructure and magnetic properties of Mn–Al–C alloy powders prepared by ball milling, Journal of Alloys and Com-pounds 622 (2015) 524-528.
34. J. Cui, M. Kramer, L. Zhou, F. Liu, A. Gabay, G. Hadjipanayis, B. Balasubra-manian, D. Sellmyer, Current progress and future challenges in rare-earth-free permanent magnets, Acta Materialia 158 (2018) 118-137.
35. T.B. Massalski, H. Okamoto, P. Subramanian, L. Kacprzak, W.W. Scott, Binary alloy phase diagrams, American Society for Metals, Metals Park, OH1986.
36. W. Dreizler, A. Menth, Transformation kinetics of the ferromagnetic alloy Mn-Al-C, IEEE Transactions on Magnetics 16(3) (1980) 534-536.
56
37. J.-O. Andersson, T. Helander, L. Höglund, P. Shi, B. Sundman, Thermo-Calc & DICTRA, computational tools for materials science, Calphad 26(2) (2002) 273-312.
38. T. Chen, W.E. Stutius, The phase transformation and physical properties of the MnBi and Mn1.08Bi compounds, IEEE Transactions on Magnetics 10(3) (1974) 581-586.
39. T.B. Massalski, Binary alloy phase diagrams, 3 (1992) 2874. 40. B. Carlsson, M. Gölin, S. Rundqvist, Determination of the homogeneity range
and refinement of the crystal structure of Fe2P, Journal of Solid State Chemistry 8(1) (1973) 57-67.
41. G.H. Gebresenbut, M.S. Andersson, P. Nordblad, M. Sahlberg, C. Pay Gómez, Tailoring Magnetic Behavior in the Tb-Au-Si Quasicrystal Approximant Sys-tem, Inorganic Chemistry 55(5) (2016) 2001-2008.
42. G.H. Gebresenbut, M. Andersson, P. Nordblad, N. Qureshi, M. Sahlberg, C. Pay Gómez, Single crystal growth, structure determination and magnetic behav-ior of RE-Au-Si quasicrystal approximants (RE= Ho and Tb), Acta Universitatis Upsaliensis 1651-6214 (2016) 30-31. .
43. Y. Jia, Y. Wu, S. Zhao, J. Wang, C. Jiang, Relation between solidification mi-crostructure and coercivity in MnAl permanent-magnet alloys, Intermetallics 96 (2018) 41-48.
44. J. Fidler, T. Schrefl, S. Hoefinger, M. Hajduga, Recent developments in hard magnetic bulk materials, Journal of Physics: Condensed Matter 16(5) (2004) S455.
45. V.K. Pecharsky, P.Y. Zavalij, Fundamentals of powder diffraction and structur-al characterization of materials, Springer 2009.
46. R.J. Tilley, Crystals and crystal structures, John Wiley & Sons 2006. 47. E. Wilson, Fifty years of synchrotrons, Proceedings of the 1996 European Parti-
cle Accelerator Conference (EPAC’96), 1996, pp. 135-139. 48. H. Rietveld, A profile refinement method for nuclear and magnetic structures,
Journal of Applied Crystallography 2(2) (1969) 65-71. 49. X. Guo, A. Zaluska, Z. Altounian, J. Ström-Olsen, The formation of single-
phase equiatomic MnBi by rapid solidification, Journal of Materials Research 5(11) (1990) 2646-2651.
50. J.B. Yang, K. Kamaraju, W.B. Yelon, W.J. James, Q. Cai, A. Bollero, Magnetic properties of the MnBi intermetallic compound, Applied Physics Letters 79(12) (2001) 1846-1848.
51. C. Lou, Q. Wang, T. Liu, N. Wei, C. Wang, J. He, Effects of a high magnetic field on the coarsening of MnBi grains solidified from isothermal annealed semi-solid melt, Journal of Alloys and Compounds 505(1) (2010) 96-100.
52. Y. Mitsui, R.Y. Umetsu, K. Koyama, K. Watanabe, Magnetic-field-induced enhancement for synthesizing ferromagnetic MnBi phase by solid-state reaction sintering, Journal of Alloys and Compounds 615(Supplement C) (2014) 131-134.
53. Y. Liu, J. Zhang, S. Cao, X. Zhang, G. Jia, Z. Ren, X. Li, C. Jing, K. Deng, Microstructure, crystallization, and magnetization behaviors in MnBi−Bi com-posites aligned by applied magnetic field, Physical Review B 72(21) (2005) 214410.
54. K. Koyama, Y. Mitsui, E.S. Choi, Y. Ikehara, E.C. Palm, K. Watanabe, Change of the equilibrium state of ferromagnetic MnBi by high magnetic fields, Journal of Alloys and Compounds 509(5) (2011) L78-L80.
57
55. Y. Sakka, M. Nakamura, K. Hoshimoto, Rapid quenching and properties of hard magnetic materials in MnAl-X (X=Ti, Cu, Ni, C, B) systems, Journal of Materials Science 24(12) (1989) 4331-4338.
56. K.P. Su, X.X. Chen, H.O. Wang, D.X. Huo, Z.W. Liu, Effect of milling on the structure and magnetic properties in Mn54Al46 flakes prepared by surfactant-assisted ball milling, Materials Characterization 114 (2016) 263-266.
57. V. Øygarden, J. Rial, A. Bollero, S. Deledda, Phase-pure τ-MnAlC produced by mechanical alloying and a one-step annealing route, Journal of Alloys and Compounds 779 (2019) 776-783.
58. K.P. Su, H.O. Wang, S. Huang, X.X. Chen, J.J. Liu, D.X. Huo, L. Ma, Z.W. Liu, Magnetic anisotropy and enhanced remanence in textured polycrystalline MnAlCuC-based flakes, Journal of Materials Science 53(13) (2018) 9823-9829.
Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1782
Editor: The Dean of the Faculty of Science and Technology
A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)
Distribution: publications.uu.seurn:nbn:se:uu:diva-379177
ACTAUNIVERSITATIS
UPSALIENSISUPPSALA
2019
AbstractList of PapersMy contribution to the papersContentsAbbreviations and units1. Introduction1.1 Background1.2 Permanent magnets1.3 Permanent magnet motors
2. Strategy and aim of the research2.1 General aspects2.2 The MnAl and MnBi phases
3. Experimental methods3.1 Synthesis methods3.1.1 Drop synthesis3.1.2 Self-flux3.1.3 Ball milling3.1.4 Cryo milling3.1.5 Heat treatment
3.2 Characterization methods3.2.1 Diffraction equipment3.2.2 Bragg’s law3.2.3 The structure factor3.2.4 Synchrotron radiation3.2.5 Neutron diffraction3.2.6 The Rietveld method3.2.7 Magnetic properties characterization3.2.8 Microstructure3.2.9 SEM3.2.10 EDS3.2.11 EBSD3.2.12 Thermal analysis
4. Results and discussion4.1 X-ray analysis of the synthesized MnAl and MnBi samples4.2 Phase stabilities of MnAl and MnBi4.2.1 The influence of cooling rate on the phase formation4.2.2 Phase stabilities and doping
4.3 Milling4.4 Microstructural analysis4.5 Magnetic properties
5. ConclusionsSvensk populärvetenskaplig sammanfattningAcknowledgementsReferences
