Essi Ristiranta
MAGNETIC PROPERTIES OF AMORPHOUS FE-SI-B GLASS-COATED MICROWIRES
Master’s Thesis
Examiners: Professor Erkki Lähderanta Junior researcher Egor Fadeev :
Lappeenranta-Lahti University of Technology LUT School of Engineering Science
Computational Engineering and Technical Physics Technical physics
Essi Ristiranta
Magnetic properties of amorphous Fe-Si-B glass-coated microwires
Master’s Thesis 2021
66 pages, 34 figures, 3 tables.
Examiners: Professor Erkki Lähderanta Junior researcher Egor Fadeev
Keywords: ferromagnet, microwires, SQUID magnetometry
Microwires are fine wires with diameter in range ofµm. These wires have attracted in- terest due to their outstanding magnetic properties, such as magnetic bistability. These properties makes them prospective candidates for different applications, e.g. sensing el- ements for various sensors. In this thesis magnetic properties were investigated for these amorphous glass-coated Fe-Si-B microwires. SQUID magnetometer was used for mag- netic measurements. Main objectives of the study were to evaluate the effect of tem- perature, nucleus diameter, thickness of the glass coating and removal of the coating on the magnetic properties of the microwires. It was observed that before glass coating re- moval all microwires exhibited magnetic bistability at all measured temperatures. After the glass coating removal microwires lost their bistable state due to partial reduction of in- ternal stresses. Exception for this was microwire with the biggest nucleus diameter, which preserved bistable state. Level of saturation for all samples decreased with temperature decrease. Reduction of nucleus diameter resulted in the increase of coercivity. Decrease of relative thickness of glass coating resulted decrease in coercivity.
Amorfisten Fe-Si-B lasipäällysteisten mikrolankojen magneettiset ominaisuudet
Diplomityö 2021
66 sivua, 34 kuvaa, 3 taulukkoa.
Tarkastajat: Professori Erkki Lähderanta Nuorempi tutkija Egor Fadeev
Hakusanat: ferromagneetti, mikrolangat, SQUID magnetometri
Mikrolangat ovat ohuita lankoja, joiden halkasija on mikrometrien luokkaa. Nämä langat ovat kiinnostavia niiden magneettisten ominaisuuksien vuoksi, kuten esimerkiksi magnet- tinen bistabiliteetti. Näiden ominaisuuksien ansiosta ne ovat potentiaalisia materiaaleja erilaisiin sovelluksiin, joista eniten kiinnostusta on herättänyt käyttö antureissa erilais- ten parametrien havaitsijana. Tässä työssä tutkittiin amorfisten Fe-Si-B lasipäällysteisten mikrolankojen magnettiisia ominaisuuksia. Mittaukset suoritettiin käyttämällä SQUID magnetometria. Tavoitteena oli tutkia lämpötilan, langan metalliosan halkaisijan, lasi- päällysteen paksuuden ja sen poistamisen vaikutuksia. Ennen päällysteen poistoa kaikki tutkitut mikrolangat olivat bistabiilisia. Päällysteen poiston jälkeen mikrolangat menet- tivät bistabiilisen tilansa lukuunottamatta mikrolankaa, jolla oli suurin metalliosan hal- kaisija. Magneettinen saturaatio heikkeni lämpötilan laskun myötä. Metallisen sisuksen halkaisijan pieneneminen lisäsi koersiviteettia. Suhteellisen lasipäällysteen paksuuden lisääminen johti koersiviteetin vähenemiseen.
I would like to thank my supervisors Erkki Lähderanta and Egor Fadeev for making this thesis possible and guiding me through this work. I would like to thank also everyone who had any contributions or are somehow related to this work.
I wish to thank my family for giving me their continuous support and friends who have made my time at LUT so special. Special thank to a friend who spent a lot of time on a call with me while we were both working with our thesis. They gave me company and support during these challenging times due to Covid-19 pandemic. There has been some great moments and some challenging ones but now it is time to head towards new challenges and see what the world has to offer.
Lappeenranta, July 15, 2021
Essi Ristiranta
2.1 Magnetic measurement units . . . 10
2.2 Paramagnetism . . . 11
2.3 Diamagnetism . . . 12
2.4 Ferromagnetism . . . 13
2.5 Antiferromagnetism . . . 14
2.6 Magnetic hysteresis . . . 15
2.7 Demagnetization . . . 16
3 MICROWIRES 18 3.1 Properties of microwires . . . 18
3.1.1 Magnetic domain structure . . . 18
3.1.2 Chemical properties . . . 20
3.1.3 Mechanical properties . . . 21
3.1.4 Electrical properties . . . 22
3.2 Prominent effects of microwires . . . 22
3.2.1 Giant magnetoimpedance (GMI) . . . 22
3.2.2 Large Barkhausen effect and magnetic bistability . . . 23
3.3 Fabrication methods . . . 23
3.3.1 Melt spinning . . . 24
3.3.2 In-rotating-water spinning . . . 24
3.3.3 Taylor method . . . 25
3.3.4 Glass-coated melt spinning . . . 26
3.3.5 Melt extraction . . . 27
3.3.6 Electrodeposition . . . 27
3.4 Applications . . . 28
3.4.1 Sensing applications . . . 28
3.4.2 Biomedical applications . . . 29
3.4.3 Other applications . . . 29
4 SQUID MAGNETOMETER 31 4.1 Working principle . . . 32
4.2 Structure of the SQUID magnetometer . . . 33
4.3 The recondensing cryostat and insert system . . . 34
4.4 The magnet . . . 34
4.5 Temperature control of the VTI . . . 35
4.6 Superconducting detection system . . . 36
4.7 Electronic rack . . . 37
5 EXPERIMENTS 39 5.1 Description of experiments . . . 39
6 RESULTS AND DISCUSSION 40 6.1 Sample 1 . . . 40
6.2 Sample 2 . . . 45
6.3 Sample 3 . . . 49
6.4 Comparison between the samples . . . 53
6.5 Current study . . . 57
6.6 Future work . . . 57
7 CONCLUSION 58
REFERENCES 59
Tables 63
Figures 64
B Magnetic flux density C Curie constant
d Nucleus diameter of wire
D Wire diameter
DC Direct current
emu Unit of magnetic moment Fe-Si-B Iron-Silicon-Boron GMI Giant magnetoimpedance H Magnetic field strength H∗ Switching field
Hc Coercive field
m Magnetic moment
M Magnetization
Mr Remanent magnetization Ms Saturation magnetization
SQUID Superconducting quantum interference device t Time
T Temperature
TC Curie temperature TN Néel temperature
VTI Variable temperature insert
1 INTRODUCTION
1.1 Background
Microwires are thin wires with diameter in the range of micrometers. The microwires can be either bulk wires or consist of an inner core and a glass coating which is shielding the core. They possess different mechanical, chemical, electrical and magnetic prop- erties. Especially magnetic microwires posses magnetic properties, like giant magne- toimpedance (GMI) effect and magnetic bistability, which allows highspeed propagation of domain wall through the entire microwire. These magnetic properties are desirable for applications such as sensing elements in various sensors. They can be used to detect different parameters, for example, temperature, tensile stress or magnetic field. Due to their small size they are very sensitive to environmental changes.
Amorphous state is favourable for GMI effect and magnetic bistability. It is known that GMI effect is strong in Co-rich amorphous microwires and magnetic bistability is inherent for amorphous glass-coated wires with positive magnetostriction. In order to fabricate microwires with these magnetic properties they are drawn together with molten glass and then quenched rapidly to obtain amorphous state. This technique was developed in sixties [1].
1.2 Objectives and delimitations
The aim of this thesis is to investigate magnetic properties of amorphous Fe-Si-B glass- coated microwires. The objectives of this thesis are:
• Since changes in magnetic domain structure affect magnetic properties, it is nec- essary to estimate influence of glass coating removal to magnetic domain structure and therefore, to magnetic properties by measuring dependencies of magnetization on magnetic field (M(H)).
• To investigate influence of the temperature on magnetic properties.
• To investigate how changes in metallic core diameter and glass coating thickness affect magnetic properties of the microwires.
measurement units and description of demagnetizing process. Chapter 3 describes mi- crowires, their properties, fabrication methods and possible application. Chapter 4 gives an overview of SQUID magnetometer, which is used to conduct the measurements in this thesis, and its working principle. Descriptions of experiments is presented in Chapter 5.
Results and discussion are provided in Chapter 6. The last Chapter 7 contains conclusions of the thesis.
2 Fundamentals of magnetism
Since the discovery of magnetism and magnetic materials in the ancient world, humankind has always been baffled and fascinated by its mysterious properties. Nowadays magnets are widely used in different areas ranging from industry to medicine. Magnets are used in everyday life in variety of electric devices such as household appliances [3, 4]. There are five main types of magnetic ordering: paramagnetism, diamagnetism, ferromagnetism, ferrimagnetism and antiferromagnetism. All materials exhibit some type of magnetic ordering [5].
The phenomenon of magnetism can not be fully explained using classical physics, also quantum mechanical approach is needed in order to understand fundamentals of mag- netism. The origin of magnetism arises from orbital angular momentum of electrons and electrons spin angular momentum [6].
2.1 Magnetic measurement units
The International System of Units (SI) is widely accepted by most nations. It is an agree- ment on which units are used as a standard. However, both SI and cgs systems are utilized for magnetic measurements. The cgs system stands for "centimeter-gram-second". In cgs system emu is an unit for magnetic moment. Magnetization M is defined as a vector sum of magnetic moments by volume, mass or moles of the sample as a counting base.
Magnetic quantities with respective units in cgs and SI systems are represented in Table 1.
Table 1.Magnetic quantities with cgs and SI units [5].
Quantity Symbol cgs units SI units
Magnetic flux density B G (gauss) T
Magnetic field strength H Oe (oersted) A m
Magnetic moment m emu Am2, J
T
Magnetization M emu
cm3 , emu g , emu
mol A m , Am2
kg , Am2 mol Magnetic susceptibility χ 1, cm3
g , cm3
mol 1, m3
g , m3 mol
One gauss corresponds to 10−4 T and one oersted corresponds to 103 4π
A
m. Gauss and
B =µ0(H+M), (2)
where µ0 represents permeability of free space. Measuring a direct magnetization re- sponse to the applied magnetic field gives
χ= M
H, (3)
where χ represents magnetic susceptibility. Magnetic susceptibility plays an important role on classifying materials [5].
2.2 Paramagnetism
Every material posses some value of magnetic susceptibility χ. Paramagnetic materials have positive values of susceptibility (χ >0). Paramagnetic susceptibility values are rel- atively small, commonly in the range of10−6 to10−1. The total magnetization depends on applied magnetic field and temperature(M ∝ BT). The paramagnetic material which is not exposed to the external magnetic field has randomly oriented intrinsic magnetic moments due to thermal agitations. Illustration of magnetic moments of paramagnetic or- dering is shown in Figure 1a. When no magnetic field is applied, these magnetic moments cancel each other out and the net magnetization is zero. When the material is exposed to an external magnetic field, the magnetic moments begin to align parallel to the field in order to decrease the Zeeman energy. However the Zeeman energy is small compared to thermal energy, hence the total magnetization is very small. Positive χvalues indicates the direction of the moments to be same as the field direction [5, 7].
Theχof many paramagnetic materials follows the equation
(a) paramgagnetism (b) ferromagnetism (c) antiferromagnetism
(d) ferrimagnetism (e) canting ferromag- netism
Figure 1.Illustration of magnetic moments of different types of magnetic ordering [5].
χ= C
T , (4)
whereCrepresents the Curie constant. The equation is known as the Curie law. The Curie law does not apply for metallic paramagnets as their susceptibility is not temperature dependent [4]. These materials close to temperature independent type of paramagnetic susceptibility are referred as materials that exhibit Pauli paramagnetism [6]. Examples of metallic paramagnets are palladium and platinum. Paramagnetic materials can be solids, liquids or gases. For example, elements sodium and oxygen are paramagnets [8].
2.3 Diamagnetism
The material is diamagnetic, when there is no innately existing magnetic moments in the material. Diamagnetic materials have negative magnetic susceptibility (χ < 0) [5].
The susceptibility values are commonly relatively small, usually in the range of−10−6 to −10−2 [7]. When an external magnetic field is applied to the diamagnetic material, it induces magnetic dipoles, which are aligned to the opposite direction of the field due to Lenz’s rule. The magnetic field penetrates into diamagnetic materials although the material tends to repel it as illustrated in Figure 2a [5].
(a) diamagnet (b) perfect diamagnet = super- conductor
Figure 2. Illustration of magnetic flux inside of diamagnets [5].
2.4 Ferromagnetism
Ferromagnetism is the form of magnetism that people normally consider as a magnet in our everyday life. Iron is a good and well-known example of the ferromagnetic material.
Materials which have similar magnetic properties to iron such as nickel and cobalt are also ferromagnets. In addition, alloys of aforementioned elements tend to exhibit ferro- magnetic behaviour [8]. Moreover, alloys with no ferromagnets as constituent elements can possess ferromagnetic behavior [9].
Ferromagnetic materials have a critical value of temperature known as the Curie temper- atureTC. BelowTC material behaves like ferromagnet with a spontaneous magnetization and above the TC material exhibits paramagnetic behaviour. The Curie law was intro- duced above in Section 2.2 describing paramagnetism, however, it does not take into account any of the magnetic moment interactions. For this reason Curie-Weiss law is introduced
χ= C
T −θ, (5)
where θ represents a correction term and is called the Weiss constant. The unit of θ is Kelvin (SI system). Positive values of θ indicates that magnetic moments have a high probability to align parallel to the external magnetic field. The alignment of the magnetic moment is illustrated in Figure 1b [5].
The origin of ferromagnetism arises from the spins of electrons [6]. The spontaneous magnetization is possible in the temperature range of 0 K toTC [3]. Ferromagnetic mate- rials exhibit magnetic hysteresis, which describes dependence of magnetization versus the applied magnetic field [4]. The hysteresis and magnetization of ferromagnets is described more in details in Section 2.6.
Susceptibility values of ferromagnets are commonly in the range of10to107. Great val- ues ofχindicates that the total magnetization of ferromagnets in the presence of magnetic field is large [7]. The susceptibility is theoretically infinite at TC and drops to the range of paramagnets above TC [3].
2.5 Antiferromagnetism
Antiferromagnets have two magnetic sublattices with magnetic moments oriented antipar- allel to each other in the presence of an external magnetic field alongside of the spin axis.
Illustration of antiparallel ordering of the magnetic sublattices is shown in Figure 1c. The magnetic moments of sublatices are equal, therefore, they cancel each others out leading to only weak net magnetism. The case of nonequivalent antiparallel magnetic sublattices is called ferrimagnetism and is presented in Figure 1d [3]. If an external magnetic field is applied perpendicular to the spin axis, the magnetic moments starts to turn with the field in the way, which is illustrated in Figure 1e. Both ferrimagnetism and canted ferromag- netism are special cases of antiferromagnetism [5, 6]. In the absence of the magnetic field antiferromagnetic material has a zero net magnetization [7].
The susceptibility range of antiferromagnets is similar to that of paramagnets [7]. The difference between them is that antiferromagnets possess ordered structure of magnetic moments. The ordered structure starts to deteriorate with the temperature increase, lead- ing to rise of susceptibility. Like ferromagnets, antiferromagnets also have a critical tem- perature. This temperature is called The Néel temperature, TN. Above TN the ordered structure is completely destroyed and the material behaves like paramagnet. The suscep- tibility of antiferromagnets reaches its peak atTN and starts to decrease as the material’s state changes to paramagnetic [6].
Figure 3. Structure of a synthetic antiferromagnet [4].
2.6 Magnetic hysteresis
Ferromagnetic materials exhibit magnetic hysteresis. Ferromagnetic materials have an irreversible nonlinear response to the applied magnetic field, and as a result they form a hysteresis loop of magnetization versus the applied magnetic field as shown in Figure 4. If a ferromagnetic material has not been exposed to a magnetic field, its magnetic domains are oriented randomly. In this case the material is in so-called virgin state. When magnetic field is applied, the magnetic domains starts to align with the field. When the magnetic domains have aligned parallel to the applied field, the material reaches a level of saturation magnetization, which is referred in Figure 4 asMs. When the magnetic field is reduced back to zero, the material will remain in the magnetized state. In Figure 4 The remanent magnetization is denoted asMr. In order to have zero magnetization in the material, it is needed to apply an opposite magnetic field of the value, which is denoted asHcin Figure 5. This value of magnetic field is called coercive field.
Figure 4. Illustration of a ferromagnetic hysteresis loop [4].
Broad loops shaped like square are characterized as hard magnetic materials. These ma- terials remain in a magnetic state even after the field is removed once they are magnetized toMsin a magnetic field. They have high coercive field values, meaning the magnetized state is not sensitive to changes in ambient fields. Therefore these materials are suitable for permanent magnets, which are used, for example, in electric motors. Materials with narrow hysteresis loops shaped like letter S are soft magnetic materials. These materials have small coercive fields. As a result, in these materials it is easy to change the sign of magnetization [4, 8].
2.7 Demagnetization
As mentioned in Section 2.6, ferromagnetic materials exhibit magnetic hysteresis. The residual magnetization in once magnetized ferromagnets can be eliminated using demag- netization methods also referred as degaussing [11]. For some applications it is necessary to demagnetize the material, for example, magnetic recording is dependent on the demag- netization.
Material can be demagnetized by exposing the material to alternating magnetic fields. The amplitude of the alternating magnetic field is gradually decreased down to zero as shown in Figure 5. As the magnetic field is decreasing, the magnetization reduces with it to its minimum value. This demagnetization method is based on the hysteresis behaviour [11].
Demagnetization can be attained using other methods as well. For example, annealing
Figure 5.Illustration of hysteresis loops during demagnetization process [11].
the material above itsTC changes its magnetic state to paramagnetic, and material loses its magnetization. After the annealing material should be cooled down in a zero magnetic field [12]. For a very rapid demagnetization effective lasers can be used to induce an ultra-fast demagnetization [13].
3 MICROWIRES
Microwires are very fine wires with diameter in the range of micrometers. Commonly, the core consist of a metal alloy, which is coated with thin layer of glass. Microwires are barely visible for a human eye, very thin, pliable, and resembling hairs.
Microwires are promising materials for a great variety of different applications, for ex- ample, in the field of sensors and biomedicine. Due to variety of properties and pos- sible applications especially ferromagnetic microwires have gained ever increasing in- terest among the scientist and companies all over the world during the last couple of decades [14].
3.1 Properties of microwires
Microwires have promising mechanical, electrical, chemical and magnetic properties.
This section describes amorphous magnetic microwires, which is the main focus of this thesis.
3.1.1 Magnetic domain structure
The magnetic domain structure of rapidly quenched wires is interesting and complex. The quenching rates differs between the centre and the surface of the wire leading to complex internal stress distribution, which usually have axial, radial and circular components. The domain structure is determined by the magnetostriction constant,λ. The magnetostriction can have positive or negative values, which is an important parameter to consider when determining the domain structure.
Amorphous wires with positive values ofλhave commonly longitudinal easy axis in the inner core and radial easy axis in the outer shell, like shown in Figure 6. The cylindrical inner core with axial domain structure covers usually around 70-90 % of the volume of the wire. This kind of domain structure enables large Barkhausen jump, which is beneficial for many applications, and also shows magnetic bistability. For instance, Fe- based microwires have positive values ofλ[15, 16].
Amorphous wires with negative values of λhave commonly axial domains in the inner
Figure 6. Schematic illustration of domain structure of wire with positive magnetostriction coef- ficient [15].
core of the wire and circular in the outer shell as shown in Figure 7. As the inner cylin- drical core has similar domain structure to the wires with positive values of λ, it is also possible for wires with negative values ofλto exhibit large Barkhausen jump. However, in the case of amorphous glass-coated wire with large negative values of λ the circular outer domain shell covers practically the whole volume of the wire. For instance, Co- based microwires have negative values ofλ[15].
Figure 7. Schematic illustration of domain structure of wire with negative magnetostriction coef- ficient [15].
Amorphous wires with values of λ close to zero have very complex domain structure, and it is extremely difficult to precisely define domain structure of them. Due to that the domain structure is most often considered to be very similar to wires with negative values ofλ. For instance, wires with composition of Co-Fe-Si-B haveλvalues close to zero [15].
The domain structure of wires, which have been fabricated by rapid quenching, can be easily modified by exposing the wire to mechanical and thermal treatments. For example, the domain structure can be significantly modified by removing the glass coating, since it
has a significant effect on the internal stress distribution of the wire. For amorphous glass- coated wires with positive values ofλthe removal of the glass coating leads to growth of the outer domain shell and to decrease of inner core, nevertheless, the whole configuration does not change. This is illustrated in Figure 8. The glass coating removal reduces signif- icantly axial tensile stress, which might soften the material in terms of magnetism [15].
(a) With glass coating
(b) Glass coating removed
Figure 8. Schematic illustration of the effect of glass coating removal on the magnetic domain structure of the wire with positive value ofλ. OS denotes outer shell and IC denotes inner core [15].
For amorphous glass-coated wires with negative values of λ the removal of the glass coating leads the change of domain structure from radial to axial as illustrated in Figure 9. The shell configuration remains unchanged, however, the volume of the outer shell domain increases. It is possible to restore the domain structure, that the microwire had before the glass coating removal, by applying tensile stress [15].
3.1.2 Chemical properties
The amorphous wires have high resistance for corrosion due to their chemical and struc- tural homogeneity as there is no grain boundaries to help the corrosion. Corrosion resis- tance even higher than in stainless steel can be obtained in amorphous Fe-based wires with addition of chromium. In addition, for glass-coated wires the coating provides protection against the corrosion [16].
Figure 9. Schematic illustration of the effect of glass coating removal on the magnetic domain structure of the wire with negative value ofλ. OS denotes outer shell and IC denotes inner core [15].
3.1.3 Mechanical properties
Amorphous microwires are found to be very strong and ductile. Some of their mechanical properties are stronger than those of their crystalline counterparts. They have high values of fracture strength approaching the theoretical values. For instance, Fe-based amorphous wires have fracture strength values up to 3.5 GPa, which is higher than of steel [15, 16].
A phenomenon called size effect is observed in glass-coated wires. The size effect means that tensile strength and thermal coefficients of glass coating and inner core alloy are connected. Meaning the wire diameter has an influence on the tensile strength. The size effect is not observed if the thermal expansion coefficients of the glass and alloy are in the same range meaning changes in nucleus diameter does not have an impact on the tensile strength. However, if the thermal expansion coefficients of glass and alloy differ significantly from each other, the size effect is observed. As the diameter decreases, the value of the tensile strength increases. The size effect was observed not only in amorphous glass coated wires, but also in crystalline glass coated wires [16].
Annealing of the wires can cause some notable changes in their mechanical properties.
For instance, Fe-based amorphous wires exhibit thermal embrittlement. The embrittle- ment can be a significant problem for some applications. Annealing can also lead to generation of nanocrystallites. On the contrary, cold-drawing can also be used to modify the mechanical properties of the wires. For instance, plasticity and tensile strength can be modified by the cold-drawing process. It can also improve a giant magnetoimpedance (GMI) effect in wires, which is desirable for sensors [15].
3.1.4 Electrical properties
Amorphous metallic wires have significantly higher electrical resistance compared to their crystalline counterparts with similar chemical composition. For instance, Fe-Si-B amor- phous glass-coated microwire has 20 % higher electrical resistance than its crystalline counterpart. In addition, the resistance is less temperature dependent. High electrical resistance is favorable for high frequency applications, but it is not for the GMI effect.
The resistance can be lowered by annealing the wire up to point of nanocrystallization.
The resistance can also be modified by external magnetic fields, by changing the chemical composition, by increasing pressure or by applying mechanical stress [15, 16].
3.2 Prominent effects of microwires
Microwires exhibit many interesting phenomena. This makes them very prospective ma- terials for a great variety of different applications in all walks of life and industry. This section presents brief overview of two of the effects.
3.2.1 Giant magnetoimpedance (GMI)
The giant magnetoimpedance (GMI) is a fairly recent discovery. The GMI effect is change of AC impedance of a soft ferromagnetic conductor under the influence of an applied mag- netic field. For a cylindrical object (e.g., microwires) the GMI effect can be understood as the result of an increase of skin depth up to the radius of the wire by means of reduction of circumferential permeability. Choosing a material with high values of circumferential permeability leads to reduced skin depths resulting in higher values of GMI, which is beneficial for sensor applications. The magnitude of the GMI effect can be modified by changing frequency.
The GMI effect can be observed in both glass-covered and conventional wires. However, for example, wire with value ofλclose to zero shows a significant decrease of the GMI value after the glass coating removal. The glass coating removal decreases the value of the circumferential permeability. In addition, glass coating thickness affects the value of GMI. Observed GMI value decreases, as the thickness of the glass coating increases.
Amorphous wires with positive values ofλ do not exhibit GMI effect. However, if the wire is annealed up to the point of nanocrystallization, it starts to exhibit the GMI effect
Materials exhibit the Barkhausen effect due to magnetization reversal in the core at a certain value of axial magnetic field, which is called the switching field (H∗). The effect is visible in hysteresis loop as a vertical step (See Figure 10). In Figure 10 the left loop represents the full effect and then the completeness of the effect decreases towards the right side. In the last hysteresis loop on the right side of Figure 10 the effect is completely vanished [18, 19].
Figure 10.Degree of the large Barkhausen effect from perfect to nonexistent [18].
The vertical jump in the hysteresis loop is also described as magnetic bistability, meaning that the magnetization switches from one stable state to another at critical value of the field. As mentioned in Chapter 3.1.1, the inner core of amorphous glass-coated wires with positive values ofλoccupies large percentage of the volume therefore the magneti- zation of the core dominates the magnetization of the wire. The inner core has two stable magnetization states which are parallel and antiparallel to the wire [18, 19].
3.3 Fabrication methods
There are several methods for microwire fabrication. This chapter introduces the most common fabrication methods: melt spinning, in-rotating-water spinning, Taylor method, glass-coated melt spinning, melt extraction and electrodeposition. Metallic amorphous
wires without glass coating are fabricated using the first two methods. Glass-coated amor- phous wires can be fabricated using Taylor or glass-coated melt spinning methods. The amorphous glass-coated microwires are the most promising for technical applications e.g.
sensing elements in sensors. Wires fabricated using electrodeposition also attract some interest among scientists due to uniform magnetic properties.
Sometimes it is needed to remove the glass coating. This usually causes changes in mag- netic and mechanical properties of a wire. The removal can be done mechanically or using chemical etching. Chemical etching usually causes less changes in magnetic, electrical and mechanical properties compared to mechanical glass coating removal. However, one should gradually decrease concentration of acid and rinse the metal with water in order to avoid etching of the metal part [20].
3.3.1 Melt spinning
The melt spinning method is widely used to fabricate amorphous metallic alloys. The cooling rate of this method ranges from 104 to 106 K/s. Using this method it is possi- ble to fabricate wires with a diameter ranging from 1 to 300 µm. The method is based on the pressure ejection of melt stream through an outlet into a coolant, after which this stream solidifies rapidly before disintegrating into droplets. In order to fabricate the wires directly from the melt with rapid solidification several conditions must be met. First, the melt stream should solidify at high cooling rates within the certain "stability" dis- tance from the outlet. Second, the coolant must have low viscosity and surface tension.
Third, the coolant must have non-turbulent and stable flow with high velocities. However, in practice it is difficult to maintain simultaneously cooling capacity of the melt stream without the precipitation of crystalline phases in the range from melting to glass transition temperature. Due to these difficulties it is challenging to to fabricate glass-coated metallic alloy wires using the melt spinning method [20].
3.3.2 In-rotating-water spinning
To overcome the limitations of the melt spinning method, the in-rotating-water spinning method was developed. The main difference is that the melt stream is guided straight into rotating water instead of impinging on the inner parts of the rotating drum. The device based on this method is presented in Figure 11. Some adjustments are needed depending on the alloy being cast. For instance, it may be necessary to adjust the ejection angle, the
Figure 11. A cross-section illustration of microwire fabricating device using in-rotating-water spinning method [20].
3.3.3 Taylor method
Taylor method was first introduced in 1924 by G.F. Taylor. In this method a metallic body is placed in a glass tube and then treated by induction heating. The heating causes the melting of the material. Contact with the molten metal leads to softening of the glass.
The softened glass can then be drawn. The drawn glass acts as mould for the solidifying metal core and provides uniform surface and diameter of the wire. The cooling rate of this methods varies from103 to 106 K/s. The diameter of the wires can be in the range of 2 to 100µm. A great variety of different metallic wires has been successfully fabri- cated using this method, including steels, noble metals and coppers. The challenges of
the Taylor method are finding a suitable sheath material which has a sufficient chemical inertness towards the used molten metal and finding a softening temperature consistent with the melting temperature of the metal. To avoid the contamination of the metal by the sheath, it is crucial to use a glass which is compatible with the metal in terms of melting temperature, viscosity and chemical properties [20, 21].
3.3.4 Glass-coated melt spinning
The glass-coated melt spinning method is also known as Taylor–Ulitovsky method [22].
This method is an alteration to the Taylor method allowing alloy systems with low wire forming capacity to be fabricated without any major difficulties. The schematic illustra- tion of the glass-coated melt spinning method is shown in Figure 12. The metal alloy is annealed using induction coils up to its melting point. Melting metal forms a droplet on the bottom of the Pyrex glass tube and softens the glass tube which is enveloping the droplet. This softened glass around the droplet allows the drawing of the capillary to take place. The molten metal fills up the glass capillary and solidifies rapidly in the solidification zone as it is cooled with the coolant jet [20, 23].
Figure 12. Schematic illustration of the glass-coated melt spinning method used for microwire fabrication [23].
The diameter of the metallic core obtained using the glass-coated melt spinning method is commonly in the range of 0.8 to 30µm and the thickness of the glass coating ranges from
3.3.5 Melt extraction
The melt extraction method has recently been widely used to fabricate magnetic mi- crowires. The schematic illustration of this fabrication method is presented in Figure 13. This method is based on the annealing of metal alloy up to its melting point. Al- loy rod is inserted to crucible to provide molten alloy. Then a wheel with sharp edge, spinning at high-speed, is contacted to the molten metal alloy droplet surface. As a re- sult wire is extracted and cooled rapidly from the alloy. Using melt extraction method the cooling rates are in the range from 105 to 106 K/s, which is even higher than of the Taylor-Ulitovsky method. Advantages using the melt extraction method are the high quality surface, smooth circular geometry of the wires, and significantly improved soft magnetic properties of the materials. The diameter of the fabricated wires commonly ranges from 30 to 60µm. These diameters are quite big compared to the other methods.
One of the major disadvantages of this method is that the diameter of the wires can not be precisely controlled [20].
3.3.6 Electrodeposition
Using electrodeposition fabrication method uniform wires with non-magnetic inner core and magnetic outer shell layers can be produced. The layers are obtained by moving the inner core through electrolyte baths and smoothing the surface with rotating rollers. The thickness of the layers can be varied by the deposition time when the value of the current is fixed. The diameter for the inner core is commonly around 20 µm and the magnetic layer thickness ranges from 2 to 7 µm, but the total diameter can vary from 20 to 1000 µm. The advantageous possibilities of electrodeposition method are: 1) to use wide range of materials (e.g., alloys, metals, composites), 2) to use continuous and batch processing, 3) to produce materials with different grain shape and size, and 4) to produce a product in the form of coating or bulk material. In addition, electrodeposition method can be combined with melt spinning method [20].
Figure 13. Schematic illustration of the melt extraction method used for microwire fabrication [20].
3.4 Applications
Microwires have a huge potential to be utilized in great variety of applications due to their versatile properties. Majority of the possible applications are related to some type of sensing. However, nowadays, microwires are not produced commercially in great quanti- ties due to higher cost compared to conventionally produced wires and design engineers’
lack of knowledge regarding the properties and availability of the microwires. Taylor wire fabrication method is cheaper fabrication technique compared to mechanical die-drawing methods [24].
3.4.1 Sensing applications
Magnetic microwires have useful properties for sensing elements in sensors. Magnetic microwires can be used as tensile stress, temperature or magnetic field sensors. One of the advantages of using microwires as sensors is their small size and mass [25].
For instance, amorphous glass-coated microwires with positive values of λ are great choice for sensors as they exhibit magnetic bistability. Working principle of the sen- sors is based on the switching of the direction of the magnetization from one stable state to another. Several external factors affect the switching field (e.g., mechanical stress,
field sensors, for example, can be used for navigation purposes and to monitor magnetic field which arise from electrical sources [25].
GMI effect is also widely used in sensor applications. GMI sensors can detect magnetic field, current (both ac and dc) and stress. Main advantages of GMI sensors are high sensitivity, low energy consumption and good thermal stability. The GMI sensors can be used, for example, in computer disk heads, antitheft systems and car traffic monitoring [17].
3.4.2 Biomedical applications
Magnetic glass-coated microwires are possible solutions for different biomedical appli- cations. Utilizing the wires in biomedical applications is related to sensing properties.
There is a need for biomedical sensors to improve healthcare and life science. With im- proved sensing it may, for example, be possible to detect diseases in early state allowing early treatment. The wires have a huge advantage over the conventional sensing devices as wires do not need separate elements for sensing and transmitting. The glass coating makes wires bio-compatible and due to their tiny size they will not cause any defects to the structure of implants. They are also very sensitive to detect different parameters due to their size. The wires can be used, for example, for sensing stress and temperature in muscles, spinal cord or bones [26, 27].
3.4.3 Other applications
Due to good mechanical properties of the amorphous wires, such as tensile strength, they can be used for reinforcement of composites or as material for cutting utensils. In addi- tion, it was observed that amorphous wires added in polymer matrix composites shows a clear reinforcement effect. That can be utilized, for example, in tyres to improve their sus- tainability [16]. Another example is silicons microwires, which have been used in solar
cells [28].
Figure 14. The "Cryogenic S700X" SQUID magnetometer, which has three main components (from left to right): cryostat, electronics rack and computer.
"Cryogenic S700X" SQUID magnetometer can detect10−6flux quanta variations in mag- netic flux. Magnetic field up to 7 T can be achieved using a superconducting magnet.
Temperature of the sample can be varied continuously from 1.6 K to 400 K.
There exist several operation modes, however, the measurements were conducted using only one mode, which is called extraction magnetometry. This is the most widely used mode, which records total magnetic moment of the sample. The sample is moved through the set of superconducting pick-up coils, which detect a magnetic flux. Compared to other extraction magnetometers this is a significant advantage as the pick-up coil system produces the electric DC/AC signal based on the magnetic fluxΦ, not the changes in the flux dΦ
dt . Therefore, it is not necessary to move the sample with a high speed through the pick-up coil system.
The distance of a sample moving up and down through the pick-up coils can be var- ied from 2 to 120 mm. For all of the measurements the distance was set to be 40 mm.
All of the measurements are conducted in a short-circuit mode of the superconducting solenoid. All of the measurements were carried out using direct current (DC) measure- ments, meaning DC current is used to generate magnetic field. In DC measurements the applied magnetic field is quantized and can have only strictly fixed values during a scan.
In order to detect magnetic hysteresis it is essential to have fixed values of the applied magnetic field during the scans [30].
4.1 Working principle
The SQUID magnetometer converts magnetic flux into electrical voltage. The SQUID magnetometer is based on the Josephson and field quantization effects. The SQUID mag- netometer has a superconducting ring with two very narrow pieces of insulator, called Josephson junctions, parallel to each other as shown in Figure 15 [31].
Josephson effect is a quantum tunneling effect where the superconducting current (Cooper pairs) is able to tunnel through the Josephson junctions. Electrons posses wave properties.
In the ring electrons are divided equally into two branches and when those branches meet again electrons converge. If there is no applied field they will meet on the other side of the ring without any phase difference. When magnetic field is applied to the system, it will induce superconducting current in the circuit. This circulating current will add to the bias current flowing through one of the junctions and subtract from that flowing through the other junction. As a result, the branches have no more the same current and there will
Figure 15.Schematic illustration of SQUID magnetometer working principle [31].
be some phase difference [29].
4.2 Structure of the SQUID magnetometer
The whole system can be divided into three main parts, which are the cryostat, the elec- tronic rack and the computer. The cryostat is liquid helium reservoir with variable tem- perature insert (VTI) in it. All the electronics components are stored in the rack. The magnetometer is controlled using a computer via the LabVIEW software [30].
4.3 The recondensing cryostat and insert system
The recondensing cryostat with the VTI in it is shown in Figure 16. The cryostat has an outer aluminium shell. The shell contains the liquid helium reservoir with VTI, the cryocooler and the recondensing loop system. The liquid helium reservoir has a thermal radiation shielding. Volume of the reservoir is 40 liters and it is fabricated from welded aluminium with the neck and tail sections made from glass fibre composite. The cry- ocooler is located at the top of the cryostat and it provides the cooling power for the system. The cryocooler liquefies helium gas using the recondensing loop. It is crucial to maintain low temperatures in the system in order to preserve superconducting parts in a superconducting state and prevent undesirable helium gas boil off. The cryostat also has a Mu metal shielding to prevent ambient magnetic fields like the Earth’s magnetic field penetrating into the system.
The VTI houses superconducting magnet and pick-up coils. Sample space is also in the VTI. The sample is inserted and removed via airlock, which is located at the top of the VTI. The sample is attached to a rod, which is used to lower the sample to the level where the pick-up coils are located. Stepper motor, which is also located at the top of the VTI, is used to move the rod (and the sample) vertically through the pick-up coils during the measurements [30].
4.4 The magnet
Schematic illustration of the superconducting magnet is shown in Figure 17. It can gen- erate magnetic field up to 7 T. The generated field is highly homogeneous and has low drift. The whole magnet can be divided into internal and external parts. The internal part generates the bulk of the magnetic field. The external part minimizes stray fields and it fine-tunes homogeneity of the axial magnetic field.
To change the value of the magnetic field it is needed to change the value of the current in superconducting magnet. This can be done by driving the superconducting bridge to a normal state, where the current can no longer flow without resistance. The change to the normal state can be achieved using a heater. Once the current is changed and stabilized, the system can be driven back to the superconducting state by switching off the heater.
In the superconducting state the current can flow without any losses, therefore it is not required to apply current to the system [30].
Figure 16.The structure of the recondensing cryostat [30].
4.5 Temperature control of the VTI
Schematic illustration of the VTI is shown in Figure 18. Liquid helium moves through a needle valve, which causes a very sharp drop of the pressure cooling down and vaporizing the helium. As the cold helium gas passes through a heat exchanger, it is warmed to the desired temperature before entering a sample chamber. Auxiliary heater, which is located
Figure 17.Schematic illustration of the magnets [30].
in the sample chamber, is also used to support the heating. There are two thermometers to monitor the temperature. Thermometer A is located at the heat exchanger and monitors the temperature of the helium gas entering to the sample chamber. Thermometer B is posi- tioned just above the sample and once the thermal equilibrium is reached the temperature given by the thermometer B is the same as the temperature of the sample [30].
4.6 Superconducting detection system
Superconducting pick-up coil detect magnetic flux. It is located outside of the sample chamber in the middle of the magnet as shown in Figure 17. The pick-up coil is part of a flux transformer, which is connected to input coils of the SQUID detection system.
Figure 18.Schematic illustration of the VTI [30].
Due to quantum effects the flux changes cause a voltage change. The SQUID detector is positioned above the magnet and it is shielded by a niobium can. The niobium can is shielding the SQUID detector from environmental noises and stray fields from the magnet.
The SQUID measures relative changes in magnetic flux, therefore it is needed to move the sample through the pick-up coils. This induces a screening current to flow in the flux transformer circuit, which opposes the resultant change in flux through the pick-up coil. This current is detected by the SQUID and it is proportional to the induced magnetic moment. In the SQUID electronics the output voltage is directly proportional to the signal, which is detected by the SQUID detector [30].
4.7 Electronic rack
The electronic rack, shown in Figure 14 consist of several parts, which are from top to bottom [30]:
• Temperature Controller: Controls the temperature of the VTI heat exchanger and monitors the thermometers.
• Level gauge and DC SQUID interface: Measures helium level in the liquid helium reservoir. SQUID output available via a BNC connector and displayed on the panel meter.
• Stepper motor controller: Controls the sample positioning within the pick-up coils.
• Data acquisition unit/Valve block indicator panel: Controls all digital and ana- logue outputs and inputs apart from the temperature controller and magnet power supply. Green LED of the front panel indicates that the computer is properly con- nected and red LED indicates power to the data acquisition unit. Front panel also has a schematic illustration of the helium circuit including electrically controlled valves. Red LED indicates that the valve is closed and green LED that the valve is open.
• Computer: Controls multiple electronic systems and runs the S700X software.
• Superconducting Magnet Power supply/Controller: Controls the current in the superconducting magnet.
• Mains power ON/OFF and Valve block: Houses electronically controlled valves which operate gas systems and also a pressure gauge for the VTI.
• Electronic filter unit and power supplies: Houses the electronic filtering circuits for all of the electrical services, which are connected to the insert, the power sup- plies for the VTI heaters, and SQUID/magnet detection circuit
2.
Table 2.Dimensions of the Fe75Si10B15microwire samples Sample D(µm) d(µm) L(mm) d/D
1 19.4 6.8 6.0 0.3505
2 26.8 10.8 5.4 0.4030
3 27.3 14.0 5.8 0.5128
5.1 Description of experiments
The sample was attached to a copper wire using a Teflon tape (Polytetrafluoroethylene PTFE) as shown in Figure 19. The copper wire was attached to a rod in order to place the sample in the SQUID magnetometer.
Figure 19.The sample attached to a copper wire using PTFE tape.
Magnetic hysteresis loop measurements for all three samples were performed at 10, 40, 70, 100 and 300 K. The magnetic field range was±50 mT. All samples were firstly mea- sured with glass coating and then without. The glass coating was removed mechanically by placing the sample between two microscope slides and carefully pressing the slides.
6 RESULTS AND DISCUSSION
In this chapter magnetic measurement results for each sample are introduced and dis- cussed. Comparison for all of the three samples at 10 K is presented. Possible future work is also discussed at the end of this chapter.
6.1 Sample 1
Hysteresis loops of sample 1 with glass coating (d= 6.8µm,D= 19.4µm,d/D= 0.3505, L = 6.0 mm) at 10, 40, 70, 100 and 300 K are presented in Figure 20. Figure 20b is an enlargement of Figure 20a to get a closer look at the hysteresis loops. Magnetic field values given in the units of T are on the x-axis and values of magnetization given in arbitrary units (a.u.) are on the y-axis. The hysteresis loops for each temperature are normalized by dividing the measured magnetization values by the maximum value of magnetization. However, there were magnetization fluctuations in magnetic fields close to the maximum measured value. Therefore, the maximum value used to normalize the data was taken just before the noisy parts.
The hysteresis loops at all measured temperatures have somewhat rectangular shapes. The rectangular shape indicates magnetic bistability of the wires, which originates from the specific magnetic domain structure. The unique domain structure is a result of stresses induced in the fabrication process, which in this case was the glass-coated melt spinning method [32]. The domain structure, which causes magnetic bistability is observed in materials with positive values of magnetostriction coefficients, as described in Section 3.1.1.
It can be seen from the Figure 20b that the squareness of the hysteresis loops increases as the values of temperature gets higher. The squareness is a consequence of the magne- tization reversal of the axial magnetic domain. The increase of the squareness indicates growth of axial magnetization meaning at higher temperatures the axial magnetization is larger. The saturation magnetization increases with temperature increase. The system aims to minimize magnetostatic energy, which is the origin of closure domains. In this case, the magnetostatic energy is decreasing with temperature decrease. This leads to decrease of the volume of the closure domains [32].
As can be seen from Figure 20, coercive field is independent on temperature. This also
(a) Hysteresis loops of sample 1
(b) Enlargement of hysteresis loops of sample 1
Figure 20.Hysteresis loops of sample 1 with the glass coating (d= 6.8µm,D= 19.4µm,d/D= 0.3505) at 10, 40, 70, 100 and 300 K.
indicates that preferential magnetization mechanism and domain structure are the same [32].
Figure 21 shows the same hysteresis loops of glass-coated sample 1 as in Figure 20, how- ever, here relative magnetization is normalized by the maximum value of magnetization
achieved at T = 10 K. Figure 21 shows that at 300 K the saturation takes place already at very small fields. This indicates sudden switching from one stable state to another. Sat- uration magnetic field at 10 K is much higher than at 300 K. Such behaviour is common for ferromagnetic materials.
Figure 21. Hysteresis loops of sample 1 with glass coating (d= 6.8µm,D = 19.4µm,d/D = 0.3505) at 10, 40, 70, 100 and 300 K normalized byMsat 10 K.
Hysteresis loops of sample 1 after glass coating removal at 10, 40, 70, 100 and 300 K are presented in Figure 22. Figure 22b is an enlargement of Figure 22a to get a closer look at the hysteresis loops. Hysteresis loops are normalized the same way as in Figure 20 for the same sample before glass coating removal.
After glass coating removal the shape of the loops changes significantly for all measured temperatures. As can be seen from Figure 22, the magnetic bistability vanishes after the removal of the glass coating. The disappearance of the magnetic bistability is a result of partially reduced stress caused by the removal of the glass coating. Interface between metallic nucleus and glass coating has a significant amount of thermoelastic tensile stress, which is caused due to difference in thermal expansion coefficients of metallic nucleus and glass coating. As the glass coating is removed, the stress is released to some extent changing the internal stress distribution [33]. Changes of internal stress distribution af- fects the domain structure. The volume of the axially magnetized domain decreases and the volume of periphery domains increases [32].
(a) Hysteresis loops of sample 1 without glass coating
(b) Enlargement of hysteresis loops of sample 1
Figure 22.Hysteresis loops of sample 1 after glass coating removal at 10, 40, 70, 100 and 300 K.
Coercive field becomes temperature dependent after removal of the glass coating. This indicates that the domain structure and preferential magnetization mechanism are not any- more similar. The domain wall no longer propagates easily through the sample, instead of that the magnetic moment vectors begins to rotate with the applied magnetic field.
Coercive field values increases as the temperature decreases. This may be a result from thermal expansion resulting an increase of axial magnetization or decrease of radial mag-
netization [32].
In Figure 23 hysteresis loops normalized by magnetic saturation Ms value at 10 K are presented for sample 1 after removal of the glass coating. Ms is a maximum value of magnetization, which is achieved when magnetization is no longer increasing despite of the increase of the magnetic field. Like hysteresis loops normalized by Ms 10 K for sample 1 with the glass coating in Figure 21, also loops for sample without the glass coating shows saturation drop with temperature decrease.
Figure 23. Sample 1 after removing the glass coating (d= 6.8µm) at 10, 40, 70, 100 and 300 K normalized byMsat 10 K.
In Figure 24 normalized hysteresis loops are presented for sample 1 at 10 K before and after the removal of the glass coating in order to investigate the influence of the glass coat- ing removal. The main visible difference between the loops observed in Figure 24 is the shape difference. Magnetic bistability state vanishes with the removal of the glass coating.
Hysteresis loop of sample 1 without the glass coating is broader, meaning coercivity in- creases with glass coating removal. The remanent magnetization gets considerably lower values for sample without the glass coating. Sample with the glass coating saturates at relatively low field compared to sample without glass coating and has bigger squareness coefficient (see Table 3). Saturation magnetization is higher for sample with the glass coating.
Figure 24.Sample 1 at 10 K with and without the glass coating.
6.2 Sample 2
Normalized hysteresis loops of sample 2 with glass coating (d= 10.8µm,D= 26.8µm, d/D = 0.4030, L = 5.4 mm) at 10, 40, 70, 100 and 300 K are presented in Figure 25.
Figure 25b is an enlargement of Figure 25a.
It can be observed from Figure 25a that hysteresis loops have rectangular shapes at all measurement temperatures indicating initial magnetic bistability [32]. The same be- haviour was observed with sample 1 as rectangular hysteresis loops are typical for glass- coated Fe-based microwires [34]. This is possible due to unique magnetic domain struc- ture, which is also exhibited by sample 1.
The enlarged Figure 25b gives a closer look of squareness of the loops and coercive fields.
The squareness of the loops increases with higher values of temperature meaning that the sample begins to show saturation at lower fields. Saturation magnetization increases with temperature increase. Coercive fields shows only little evidence of temperature depen- dence as the coercivity does not increase much with decrease of temperature. In fact temperatures 10 K and 40 have the same coercive field values.
Figure 26 Shows hysteresis loops of sample 2 with the glass coating for whole range of measurement temperatures normalized by Ms at 10 K. Level of saturation decreases as temperature decreases. Same behaviour was observed with sample 1. This is characteris-
(a) Hysteresis loops of sample 2 with the glass coating
(b) Enlargement of hysteresis loops
Figure 25. Hysteresis loops of sample 2 with the glass coating (d= 10.8µm,D= 26.8µm,d/D
= 0.4030) at 10, 40, 70, 100 and 300 K.
tic behaviour for ferromagnetic materials.
Hysteresis loops of sample 2 after removing the glass coating for the whole range of mea- surement temperatures are presented in Figure 27. Figure 27b is an enlargement of Figure 27a. Behaviour exhibited by sample 1, losing the rectangular shape of hysteresis loops
Figure 26.Sample 2 with glass coating (d= 10.8µm,D= 26.8µm,d/D= 0.4030) at 10, 40, 70, 100 and 300 K normalized byMsat 10 K.
after removal of the glass coating indicating disappearance of bistable state, is observed also for sample 2 after the glass coating removal. This is also a result of changes in in- ternal stress distribution caused by the partial reduction of stresses by removing the glass coating. The changes in internal stress distribution leads to changes in magnetic domain structure decreasing volume of the axial domain and increasing volume of the periphery domains.
Coercive field values are increasing with the decrease of temperature. Coercivity increases at all measurement temperatures compared to results before glass coating removal (see Table 3). This indicates that preferential mechanism of magnetization reversal changes from the high speed magnetic domain wall propagation through the wire to rotation of magnetic moment vectors with the applied external magnetic field.
Figure 28 is a comparison normalized byMsat 10 K. Saturation drops with the tempera- ture. This observation is in the line with results observed with sample 1.
Figure 29 shows hysteresis loops of sample 2 before and after glass coating removal at 10 K. Before glass coating removal sample shows magnetic bistability. The bistable state is lost after glass coating removal, due to changes in magnetic domain structure caused by partial stress reduction. More precisely, volume of axial domain decreases and periphery domains increases. Squareness coefficient is bigger for sample before glass-
(a) Hysteresis loops of sample 2 without glass coating
(b) Enlargement of hysteresis loops
Figure 27. Hysteresis loops of sample 2 after removing the glass coating at 10, 40, 70, 100 and 300 K.
coating removal. Saturation magnetization values are higher for sample with the glass coating. Coercivity increases with the glass coating removal.
Figure 28.Sample 2 after removing the glass coating (d= 10.8µm) at 10, 40, 70, 100 and 300 K normalized byMsat 10 K.
Figure 29.Sample 2 at 10 K with and without the glass coating.
6.3 Sample 3
Normalized hysteresis loops of sample 3 with glass coating (d= 14.0µm,D= 27.3µm, d/D = 0.5128, L = 5.4 mm) at 10, 40, 70, 100 and 300 K are presented in Figure 30.
Figure 30b is an enlargement of Figure 30a.
As can be observed from Figure 30a hysteresis loops of the third sample are rectangular for the whole range of measurement temperatures indicating magnetic bistability. Same observations were done with the first and second sample. This is also caused by the same unique magnetic domain structure. Remanent magnetization increases with the increase of temperature.
One can observe from the enlarged Figure 30b that coercivity differs between the tem- peratures. A little increase of coercivity is observed with increase of temperature. How- ever, all the coercive field values are very close to each other. Therefore, the preferential magnetization mechanism and domain structure can be considered to be the same. This means that magnetization reversal happens with high speed propagation of the domain wall through the entire wire.
Hysteresis loops of sample 3 after removing the glass coating for the whole range of measurement temperatures are presented in Figure 31. Figure 31b is an enlargement of Figure 31a.
The shape difference of hysteresis loops observed in Figure 31a compared to ones with glass coating presented in Figure 30a is well visible. The loops in Figure 31a are broader loops with smaller squareness coefficients (see Table 3). One key observation is that the sample preserves the bistable state after the removal of the glass coating unlike the first and second samples. This is observed most clearly at higher temperatures and bistability begins to fade away with the decrease of temperature. Bistable state indicates similar magnetic domain structure, which is exhibited by all of the samples before the glass coat- ing removal.
The squareness increases with the increase of temperature. Coercive field achieves higher values with decrease of the temperature. The coercive field also gets higher values com- pared to the same sample before the glass coating removal. This fact indicates some minor changes in magnetic domain structure. These changes are a result of the small changes in internal stress distribution, which is caused by the removal of the glass coating. The changes causes small movement of the domain wall between axially magnetized domain and periphery domains. The volume of the axially magnetized core decreases a little and periphery areas volume increases, respectively. Also the volume of the closure domains increases. This change can be considered insignificant in terms of changing the whole domain structure drastically causing the disappearance of the bistable state [32, 35]. It is worth to mention that sample 3 has the biggest nucleus diameter out of the studied samples and also the biggestd/Dratio.
(a) Hysteresis loops of sample 3 with the glass coating
(b) Enlargement of hysteresis loops
Figure 30. Hysteresis loops of sample 3 with the glass coating (d= 14.0µm,D= 27.3µm,d/D
= 0.5128) at 10, 40, 70, 100 and 300 K.
Figure 32 shows hysteresis loops of the third sample before and after glass coating re- moval at measurement temperature 10 K. The Figure 32 illustrates well the transforma- tion of the hysteresis loop from a narrow to a broader one as the glass coating is removed.
The sample before glass coating is in a bistable state. After the glass coating removal the bistable state is still preserved to some extent. The switching field is higher after the
(a) Hysteresis loops of sample 3 without glass coating
(b) Enlargement of hysteresis loops
Figure 31. Hysteresis loops of sample 3 after removing the glass coating at 10, 40, 70, 100 and 300 K.
glass coating removal. Hysteresis loop of the sample without the glass coating has lower values for remanent magnetization compared to one with the glass coating and squareness coefficient is also smaller.
Figure 32.Sample 3 at 10 K with and without the glass coating.
6.4 Comparison between the samples
Normalized hysteresis loops of all three samples with glass coating (d1 = 6.8µm, D1 = 19.4 µm, d1/D1 = 0.3505; d2 = 10.8µm, D2 = 26.8µm,d2/D2 = 0.4030; d3 = 14.0µm, D3 = 27.3 µm, d3/D3 = 0.5128) at 10 K are presented in Figure 33. Figure 33b is an enlargement of Figure 33a. Loops are very narrow in general, which indicates that the material can be classified as a soft ferromagnet. The softer the magnet is, the easier it is to change the sign of magnetization [36].
All three samples have rectangular hysteresis loops, which is common for Fe-based mi- crowires covered with glass coating. This indicates that magnetic bistability is exhibited by each wire at low temperatures, which in this case was 10 K.
The coercive field values differs slightly between the samples. The field value is the smallest for the third sample with just a small difference compared to the second sample.
The first sample has the biggest coercive field value. The first sample has the smallest nucleus diameter and in comparison the third sample has the biggest nucleus diameter.
In conclusion the coercivity increases with the decrease of the nucleus diameter. Also as the ratio of d/D increases, the coercivity decreases. The ratio ofd/D influences on the amount of internal stress, which can be seen as the consequence of different coercive field values. Increase of the ratiod/D decreases internal stresses. The broader the hysteresis loop is the harder the magnet is considered to be. In other words, the increase of the ratio
(a) Hysteresis loops of all samples with the glass coating
(b) Enlargement of hysteresis loops
Figure 33.Comparison between all of the samples with the glass coating at 10 K
d/D softens the magnet [37].
It is worth to emphasize that length also effects the properties of the wires. The effect of length is more significant for thicker wires. However, the effect of length is not considered in more detail in the current thesis.
(a) Hysteresis loops of all samples without the glass coating
(b) Enlargement of hysteresis loops
Figure 34. Comparison between all of the samples (d1 = 6.8 µm; d2 = 10.6µm;d3= 14.0µm) after the removal of the glass coating at 10 K
From the enlarged Figure 34b one can observe that coercivity is highest for sample 1 and smallest for sample 3. Coercivity is bigger for samples with smaller nucleus radius.
