Study of relaxation phenomena and local structure evolution in metallic glasses by means of Mössbauer and mechanical spectroscopy

PhD program in Computational and Applied Physics STUDY OF RELAXATION PHENOMENA AND LOCAL STRUCTURE EVOLUTION IN METALLIC GLASSES BY MEANS OF MOSSBAUER AND MECHANICAL SPECTROSCOPY Doctoral thesis by: Seyedeh Leila Panahi Thesis advisors: Pere Bruna and Eloi Pineda Department of Physics, Institute of Energy Technologies Universitat Politècnica de Catalunya Barcelona, December 2022 This thesis is dedicated to my mum. Thank you for always believing in me. Acknowledgments I want to start by saying how grateful I am to all people, my heroes, who have given me so much of their time, love and energy and supported me during my Ph.D. studies. I’d like to express my deep and sincere gratitude to my supervisors, Dr. Pere Bruna and Dr. Eloi Pineda, for their encouragement, guidance and support throughout the duration of this study and opportunity that provide for me to learn more and progress more. I would like to express my special thanks to Dr. Pere Bruna for supporting me to apply for scholarship and always be available for consult during experiment and after experiment for analyzing the data. In addition, I’d like to thank to Professor Daniel Crespo, The Rector of university who offer me so much of help and good advice during this Ph.D. and experiments. He always followed my activities with enormous patience. I’m so grateful to have opportunity to learn hard working attitude, positive energy and being humble and helper beside of all brilliant knowledge that he has in research, his brilliant way for solving problem is always inspiring me to never give up in any point of life. Also, I would like to thank professor Trinitat Pradell who teachs me how prepare samples for Nanoindentation and other measurements. I want to thank Dr. Trifon Trifonov from Center de Recerca en Nanoenginyeria for X-ray Diffraction and SEM and XPS, Special thanks go to thank Dr. Jordi Sort and Dr. Jordina Fornell from Universitat Autonomia de Barcelona (UAB) for their help and guidance for nano-indentation test. Im grateful for all brilliant people: Chenyang, Georgy, Mehran, Mahin, Neda, Lucy and David in our office for their friendship, support and kindness. Finally, My deepest gratitude goes to my family without whom I couldn’t have achieve my goals, Thank you mum for all your patient, support, and kindness and for seeing ability inside me and let me to show it, for encourage me to be brave and fighter and don’t be afraid to take new risk and challenge, I hope you see all and you know all is because of you, I would like to express my eternal love to My Dad , My dearest sister Sami and my affectionate brother Hamed, my partner Anxo Thank you for all support , energy, love and care ,without you it’s impossible finish this journey. Table of Contents List of figures .................................................................................................................... I List of tables ................................................................................................................... VI List of academic publications, activities and grants ........................................................ VII CHAPTER 1. INTRODUCTION...................................................................................... 1 1.1. Metallic Glasses ..................................................................................................... 2 1.2. Glass forming ability .............................................................................................. 4 1.2.1. Parameters to describe the glass forming ability ............................................... 5 1.3. High entropy alloys ................................................................................................ 6 1.3.1. Properties and applications of High-Entropy Alloys ......................................... 9 1.4. Alloy design ........................................................................................................... 9 CHAPTER 2. EXPERIMENTAL METHODS................................................................ 11 2.1. Sample preparation ............................................................................................... 11 2.2. X-Ray Diffraction (XRD) ..................................................................................... 13 2.3. Synchrotron X-ray diffraction............................................................................... 14 2.4. Differential Scanning Calorimetry (DSC) ............................................................. 16 2.5. Scanning Electron Microscope (SEM) .................................................................. 17 2.6. X-ray Photoelectron Spectroscopy (XPS) ............................................................. 18 2.7. Mössbauer Spectroscopy ...................................................................................... 18 2.8. Nanoindentation ................................................................................................... 19 2.9. Corrosion ............................................................................................................. 21 2.9.2. Experimental details....................................................................................... 25 CHAPTER3. MICROSTRUCTURAL CHARACTERIZATION .................................... 27 3.1 Production of (FeCoCrNi) based alloys ................................................................. 27 3.2. Structural characterization .................................................................................... 28 3.3. Glass formation .................................................................................................... 29 3.4. Thermal characterization ...................................................................................... 32 3.5. Structural characterization and effect of annealing on the as-quenched ribbons..... 34 3.6. In-situ characterization of the crystallization......................................................... 36 3.7. Characterization by Mössbauer spectroscopy........................................................ 42 3.8. Effect of annealing on the particle size distribution ............................................... 48 3.9. Discussion ............................................................................................................ 50 CHAPTER 4. MECHANICAL CHARACTERIZATION ............................................... 53 4.1. Nanoindentation ................................................................................................... 53 4.2. Discussion ............................................................................................................ 62 CHAPTER5.ELECTROCHEMICAL CHARACTERIZATION ..................................... 65 5.1. Electrochemical measurements in NaCl solution .................................................. 66 5.2. Electrochemical Impedance Spectroscopy (EIS) measurements ............................ 68 5.3. XPS analysis of A, AB20 and AB10Si10 alloys immersed in NaCl ...................... 71 5.4. Microstructure characterization after immersion in NaCl ...................................... 75 CHAPTER6. CONCLUSIONS ....................................................................................... 77 References ...................................................................................................................... 81 I List of figures Figure 1.1 Schematic ternary alloy system where the blue corner region indicates the conventional alloys based on one or two principal elements whereas the red center region indicates the high-entropy alloys. Figure 1.2 Schematic time-temperature-transformation (T-T-T) diagram. Figure 1.3 Change of volume of a molten alloy as a function of the temperature. Figure 1.4 Number of equiatomic compositions as a function of the constituent elements together with a sketch of the location in a phase diagram of the high-entropy and conventional alloys. Figure 1.5 A typical two-dimensional 𝛿−∆𝐻𝑚𝑖𝑥 plot showing the phase formation in HEAs, including the regions where solid solutions (SS) or amorphous (AM) structures can be found. Figure 2.1 (a) Arc-melter, (b) scheme of the melt spinner, (c) produced ribbons. Figure 2.2 (a) XRD pattern of a crystalline alloy and (b) XRD pattern of amorphous samples. Figure 2.3 (a) BL04-MSPD (Material science and powder diffraction Beamline setup. (b) Linkam holder with the Al disc and the attached ribbons. Figure 2.4 NETZECH 404 F3 Differential Scanning Calorimetry (DSC). Figure 2.5 Nanoindentation device in Universitat Autònoma de Barcelona with a Berkovich indenter Figure 2.6 Continuous stiffness measurement (CSM) showing: (a) the detail of the loading curve and (b) the calculation of the stiffness by the slope of unloading curve. Figure 2.7 Schematic corrosion process illustrating the anodic and cathodic current components VIII  Oral Presentation in EUROMAT 2019 Stockholm, Sweden: Influence of the alloying element on structural, mechanical, thermal and magnetic properties of a new iron base high-entropy metallic glasses  Oral Presentation in HERCULES European School, March 2020, Grenoble, France: Study of new high entropy metal glasses Other activities:  Hercules European Research Course for Users of Large Experimental Systems. Fulltime participant (100 hours of lectures and 18 hours tutorials) Grenoble, France.  Participation in synchrotron experiment: Disorder induced resonance in metallic glasses. DIAMOND, proposal ID: SM18772-1, Spring 2018, Harwell Campus, Oxfordshire.  Participation in synchrotron experiment: Investigation of the glass forming ability and in-situ crystallization of a novel family of high-entropy metallic glasses. ALBA, proposal ID: 2018022691, Autumn 2018, Cerdanyola del Vallès.  Participation in synchrotron experiment: Low temperature measurement of disordered induced ressonance in metallic glasses. ALBA, proposal ID: 2019023301, Autumn 2019, Cerdanyola del Vallès. 1 CHAPTER 1. INTRODUCTION From ancient times metals have been attractive objects for humankind and with the progress of science the need for using metals for different applications have increased drastically. For many years, researchers used traditional approaches for the discovery of new materials based on slow trial and error techniques. In this traditional way, scientists chose the elements and their proportion randomly, followed the process of production and finally tested the material to observe their properties. They checked if the material properties could be useful for any particular application, otherwise the process started from the beginning but changing the constituent elements or their percentage. In modern times, scientists tried various methods to improve the efficiency of metals and one of the best and more recent examples of this is the advent of the so-called Metallic Glasses (MGs). Metallic glasses are metallic alloys with a disordered structure that are produced when the metallic liquid is rapidly quenched to solid state bypassing crystallization. Thus, the atoms have a local disordered atomic packaging and a lack of the main features of crystalline materials such as atomic planes, crystalline directions, grain boundaries and crystalline defects [1][2]. MGs can be used in plenty of applications such as soft (e.g. transformer core laminations) or hard magnetic materials, wear-resistant light alloys, materials with enhanced catalytic performance for fuel-cell applications and new alloys for medical implants and dental amalgams [3]. On the other side, in the last decades, some groups have focused their research on a new family of alloys which are called high entropy alloys (HEAs). These alloys are attractive because of their unique composition, microstructure and adjustable properties. HEAs contain more than 5 main elements in equal or near equal atomic percentage (at%) and the atomic fraction of each component is greater than 5 at%. Thus, in a multicomponent phase diagram, HEAs are situated in the center of the diagram (see figure 1.1). 2 Figure 1.1 Schematic ternary alloy system where the blue corner region indicate the conventional alloys based on one or two principal elements whereas the red center region indicates the ‘highentropy’ region (taken from [4]). Besides of that, if we consider the number of elements which are used for the new alloys we can see that there are some elements which are normally used frequently while there are others scarcely used because of the difficulty during their processing, high prices, toxicity, etc. For high entropy alloys normally five or more principal elements are used and this creates the opportunity to use new elements from the periodic table and to investigate their new properties for different applications. 1.1. Metallic Glasses Metallic glasses were produced for the first time by Professor Pol Duwez in 1960 in the California Institute of Technology at Pasadena. He synthesized the Au75Si25 alloy by rapid solidification [1]. After that, hundreds and thousands of alloys with different compositions have been produced as metallic glasses. In the early 1990s metal-metal systems such as La-, Mg-, and Zrbased alloys were prepared by quenching from the supercooled liquid. The main difference between metallic glasses and normal metallic materials is that metallic materials are crystalline in their nature and the constituent elements are arranged in a periodic manner, while metallic glasses have a disordered atomic structure which is produced by rapid solidification from the liquid state. In many cases the cooling rate is very fast, between 105 - 106 K s-1. These high cooling rates are needed because metallic melts have a high tendency to crystallize and, therefore, high cooling rates are required to prevent crystallization. If the resultant alloy is fully 3 amorphous it is characterized by a broad diffuse halo in an X-ray diffraction pattern, instead of the showing the Bragg peaks associated to crystalline planes. In most of the cases, glassy alloys can be produced only as thin ribbons, wires or sheets. However, in 1995, the first Fe-based bulk metallic glasses (BMGs) were successfully developed in the Fe-Al-Ga-P-C-B compositional system [2]. BMGs are non-crystalline metallic alloys that can be produced with lower cooling rates than conventional MGs, between 10 and 100 K s-1 depending on the composition of the alloy. Therefore, it is possible to produce BMGs that are several centimeters thick as the cooling rate is inversely proportional to the diameter of the resultant ingot. To obtain a supercooled metallic liquid with enough glass-forming ability (GFA) to form a BMG there are three empirical rules which are known as Inoue’s rules [5]: (1) the multicomponent system should consist of three or more elements; (2) there should be a significant difference between the atomic size of such constituent elements (>12%) and (3) the elements should have a negative heat of mixing between them. During the last 20 years a large variety of metallic glass-forming compositions have been discovered and they can be classified into two big groups: metal-metal and metal-metalloid types [3] Metal-metalloid type contains around 80% of metal atoms and 20% of metalloid atoms (B, C, P, and Si). The metallic content can be of different types of metallic atoms such as, for example, in the well-known cases of Au80Si20, Fe80B20, Pd80Si20, Pd77Cu6Si17 or Fe40Ni40B20. In the case of the metal-metal type, they just contain transition metals without any particular restriction regarding the atomic percentage of the elements, such as, for example, the Fe90Zr10, Ni60Nb40, Mg70Zn30 or Cu57Zr43 binary systems. In the present work we will take as a starting point a well-known alloy of the metal-metalloid type with 20 at% of B and 80 at% of Fe that shows good soft magnetic properties, high abrasive resistance, prominent corrosion resistance and high strength. However, these Fe-based metallic glasses have the limitation of a reduced critical diameter, few millimeters when produced as bulk amorphous alloys, because of their low GFA. Therefore, there is a limitation for the potential use of Fe-based metallic glasses in engineering applications and, nowadays their production is mostly in form of thin ribbons, wires and films [6]. The first Fe-based bulk metallic glass, Fe-(Al,Ga)-(P,C,B,Ge,Si), was produced in a shape of a rod with 1-2 mm diameter in 1995 [7]. The largest thickness achieved for Fe-based metallic glass is around 16 mm whereas Pd-based metallic glasses can reach several centimeters. In many cases, researchers tried to improve the GFA of Fe-based amorphous alloys and to produce bulk Fe-based metallic glasses by adding elements like Al, Mo, Nb, Ga and Y in amounts of 1-30 at%. 4 1.2. Glass forming ability There are many different methods to produce metallic glasses, like rapid solidification, vapor deposition or electrodeposition. When an alloy is cooled down with sufficient cooling rate from the equilibrium liquid state, the metallic melt become undercooled and bypasses crystallization. However, each alloy has a different critical cooling rate and a parameter is required in order to characterize how easy is for a melt to become an amorphous solid upon cooling. The GFA is the ability of a melted alloy to solidify without crystallization in a completely amorphous structure. It can be estimated directly by the critical cooling rate (𝑅𝑐), the minimum cooling rate which is required to completely suppress crystallization. 𝑅𝑐 can be evaluated by using the Time-Temperature-Transformation (TTT) diagram. Figure 1.2 shows a schematic TTT diagram for a melted alloy which is cooled down from above the liquidus temperature (𝑇𝑙) to below the glass transition temperature (𝑇𝑔) with two different cooling rates. The glass transition temperature (𝑇𝑔), is the temperature at which the transition from supercooled liquid to an amorphous solid takes place. Figure 1.2 Schematic time-temperature-transformation (TTT) diagram (taken from ref. [3]). According to curve (2) in Figure 1.2, a glass can be produced if the cooling process is fast enough to avoid the crystallization of the melt. However, if the cooling is slower, like the one indicated with curve (1) in the figure, the result will be a crystalline solid. The glass forming 5 ability of a melt is in direct relation with the maximum thickness (𝑡𝑚𝑎𝑥) in which the alloy can be produced, being this thickness larger when the GFA is higher. 1.2.1. Parameters to describe the glass forming ability There are several parameters that can be used to describe and characterize the glass forming ability of metallic glasses, being the supercooled liquid region and the reduced glass transition temperature the most used ones. A) Supercooled liquid region (𝜟𝑻𝒙) As the temperature of a molten alloy decreases, the specific volume and the enthalpy also decrease becoming more and more viscous until it forms a glass. By convention, it is considered that a glass is formed when the viscosity reaches a value of 1012 Pa s, therefore this region is considered as the glass transition region. Figure 1.3 shows the change in volume with temperature and the glass transition temperature. The value of 𝑇𝑔 is usually determined upon heating by means of a differential scanning calorimetry (DSC) or differential thermal analysis (DTA) with heating rates of 10-20 K min-1 from the increase in the heat capacity of the alloy. A proper choice of the alloying system and its composition can reduce the critical cooling rate for glass formation. As a result, the thickness of the glass parts can increase from few micrometers (20-50 µm) in rapid solidification processes to few millimeters or few centimeters using conventional casting techniques [8]. According to Figure 1.3, in the heating process from a glassy state the alloy crosses different regions until reaching the melting point. The increase in temperature decreases the viscosity of the alloy and increases its volume until it reaches the glass transition temperature where the disordered material enters in the supercooled liquid region. A further increase in temperature leads to the melting point, 𝑇𝑚, above it the liquid is the stable thermodynamic phase. In the supercooled liquid region, the material can crystallize transforming to the stable thermodynamic ordered phase. As it can be observed in the figure there is not a uniquely defined 𝑇𝑔 but rather a glass transition range of temperatures, since the transition from a glassy state to supercooled liquid depends on the kinetics. The region between 𝑇𝑔 and the crystallization temperature (𝑇𝑥) is known as the supercooled liquid region (SLR), 𝛥𝑇𝑥= 𝑇𝑥-𝑇𝑔. 𝛥𝑇𝑥 is a parameter that indicates the thermal stability of the liquid phase and its resistance to crystallization. A larger value of 𝛥𝑇𝑥 means that the supercooled liquid can exist in a wide 6 temperature range without crystallization and that it has a high resistance to nucleation and growth of the crystalline phases [5]. Figure 1.3 Change of volume of a molten alloy as a function of the temperature (taken from ref. [9]). B) Reduced glass transition temperature (𝑻𝒓𝒈) In 1969 Turnbull, for the first time, used the reduced glass transition temperature (𝑇𝑟𝑔) as a parameter to assess the GFA of glass formers and now it has become widely used. The reduced glass transition temperature is defined as [10] 𝑇𝑟𝑔= 𝑇𝑔 𝑇𝑙 (1.1) where 𝑇𝑔 is the glass transition temperature and 𝑇𝑙 is the liquidus temperature of the alloy. Thus, the higher 𝑇𝑔 and lower 𝑇𝑙, the higher the tendency to form glass and, therefore, a large value of 𝑇𝑟𝑔 means a high GFA. The GFA of a system is not just related to the cooling rate but it also depends on the alloy composition. The 𝑇𝑟𝑔 tends to be higher for large number of alloying elements and, therefore, it is more easy to produce glassy multicomponent alloys than simple binary or monoatomic systems [11]. 1.3. High entropy alloys For many years, researchers focused on the corner of the phase diagrams to develop new alloys. However, as illustrated in Figure 1.4, the shifting of the focus to the center of the diagram gave rise to the appearance of the so-called high-entropy alloys (HEAs). 7 Figure 1.4 Number of equiatomic compositions as a function of the constituent elements together with a sketch of the location in a phase diagram of the high-entropy and conventional alloys (taken from ref. [4]). High entropy alloys were for the first time introduced by Jien-Wei Yeh and coworkers. They defined them as multicomponent solid solution alloys that contain more than five constituent elements in equiatomic, or close to equiatomic, ratios. Such condition increases the configurational entropy of mixing by an amount enough to overcome the enthalpies of compound formation [12]. The configurational entropy, 𝑆, of a random mix of 𝑁 elements is defined by the following equation ∆𝑆=−𝑅∑𝑥𝑖ln𝑥𝑖 𝑁 1 (1.2) where 𝑅 is the gas constant, 𝑥𝑖 is the mole fraction of the element 𝑖 in the alloy and ∑𝑥𝑖=1 𝑁 1. Also, the entropy of mixing for 𝑁 elements is ∆𝑆𝑚=𝑅ln𝑁 (1.3) According to equation (1.3) by increasing the number of alloying elements the entropy of mixing increases quickly. High entropy alloys can form crystalline structures like simple, single phase solid solutions with different crystallographic arrangements, e.g., face centered cubic (FCC), body centered cubic (BCC) or hexagonal close-packed (HCP) and in some cases can 8 be also arranged in an amorphous structure forming the so-called high-entropy metallic glasses (HEMGs). The structure of HEAs can be predicted mainly by two parameters [13]: the atomic size difference (𝛿) and the mixing enthalpy (𝛥𝐻𝑚𝑖𝑥). The atomic size difference is defined as 𝛿=√∑𝑐𝑖(1−𝑟ᵢ𝑟)2 𝑛 𝑖=1 (1.4) where 𝑟 is the average atomic radius and 𝑐𝑖 and 𝑟𝑖 are the atomic percentage and atomic radius of the 𝑖-th component [14]. The mixing enthalpy is defined as 𝛥𝐻𝑚𝑖𝑥= ∑ 4Ω𝑖𝑗𝑐𝑖𝑐𝑗 𝑛 𝑖=1,𝑖≠𝑗 (1.5) Here Ω𝑖𝑗=𝛥𝐻𝑚𝑖𝑥−𝑖𝑗 is a regular solution interaction parameter between the i-th and j-th elements, representing the enthalpy of mixing for binary equal atomic alloys composed of these two components. The values of 𝛿 and 𝛥𝐻𝑚𝑖𝑥 determine if a solid solution or an amorphous phase is formed (see figure 1.5). In particular, a solid solution can be produced if 𝛿 is small (𝛿 < ~0.066) and 𝛥𝐻𝑚𝑖𝑥 is either slightly positive or significantly negative (-11.6 < 𝛥𝐻𝑚𝑖𝑥 < 3.2 kJ mol-1) whereas an amorphous phase can be produced if 𝛿 is large (𝛿 ˃ 0.064), and 𝛥𝐻𝑚𝑖𝑥 is negative (𝛥𝐻𝑚𝑖𝑥 < -12.2 kJ mol−1) [8]. 9 Figure 1.5 A typical two-dimensional 𝛿−𝛥𝐻𝑚𝑖𝑥 plot showing the phase formation in HEAs, including the regions where solid solutions (SS) or amorphous (AM) structures can be found (taken from ref. [15]). 1.3.1. Properties and applications of High-Entropy Alloys HEAs have an intensive lattice distortion effect which is related to the different atomic radius of the various elements. This kind of effect can prevent dislocation movements and yield to these materials a high strength and hardness. At the same time the interaction of the different alloying elements is exacerbated, thus, high entropy alloys reflect the comprehensive properties of all kind of alloying elements. High entropy alloys also have a nice corrosion resistance, due to the simplified crystal structure, that can be higher than the one of widely used stainless steels [16]. High entropy alloys can be used as cutting tools. We can compare the friction and wear behavior of high entropy cutting tools with ordinary steels: while external load and rotational speed are fixed, the friction of high entropy alloy coated cutting tools is always lower than the ordinary high-speed steel. High entropy alloys used as coatings on the surface of cutting tools can reduce the friction force in the cutting process, extend the tool life and amend the surface quality of machined surfaces [16]. Moreover, some high entropy alloys have excellent hydrogen storage capacity. Sahlberg et al. demonstrated that a high the entropy alloy TiVZrNbHf can absorb a much higher amount of hydrogen than its constituents due to its distorted BCC-structure which has a large capacity to store hydrogen. In general a simple structure, often BCC or FCC, and the presence of lattice strains in the lattice or at interfaces can be favorable for hydrogen storage due to the variation in atomic size radii of the constituent atoms [17]. 1.4. Alloy design In the present thesis a new composition of a high-entropy metallic glass has been produced and characterized combining a well-known metallic glass system (Fe80B20) with a high-entropy alloy (FeCoCrNi). On one side, Fe80B20 has an amorphous structure at room temperature and it exhibits great soft magnetic properties, high abrasive resistance and good corrosion resistance. On the other side, FeCoCrNi is an equiatomic high entropy alloy with both high tensile strength and ductility with a face-centered cubic (FCC) structure [18]. It is a well-known strategy for improving the glass forming ability of an alloy to add elements with a low atomic weight, such 16 The ribbons with a thickness of about 20-30 m and a width of 2-3 mm were mounted on an Aluminum disc with a 3 mm hole in the middle that is attached to the Linkam. The ribbons were hold in the hot stage with silver paint (see figure 2.3b). The temperature increased from room temperature up to 873 K at a heating rate of 20 K min-1 and a spectra was recorded every 5 seconds. The diffraction pattern of the background was recorded at room temperature and subsequently subtracted from the sample signal after intensity normalization. 2.4. Differential Scanning Calorimetry (DSC) Differential Scanning calorimetry (DSC) is the most commonly used method for studying the thermal properties of various materials. In DSC, the sample and reference are kept in two places and the temperature rises in both at the same rate, so the instrument begins to measure the difference in heat flow between the sample and reference. In the constant heating mode, the temperature increases linearly with time [25]. DSC is used to measure different material properties such as the glass transition temperature 𝑇𝑔, the crystallization temperature 𝑇𝑥, the melting point 𝑇𝑚, the heat capacity, etc. DSC shows exothermic and endothermic peaks that can be associated with crystallization events and the melting point respectively. with DSC data, the reduced glass transition temperature 𝑇𝑟𝑔 and the supercooled liquid region 𝛥𝑇𝑥 can be calculated to study the thermal stability of the system. In this work, the DSC measurements of the alloys were performed using a NETZSCH 404 F3 instrument, shown in figure 2.4. The heating rate used was 20 K min-1 and from room temperature up to 1573 K under a flow of Nitrogen. 17 Figure 2.4 NETZECH 404 F3 Differential Scanning Calorimetry (DSC). 2.5. Scanning Electron Microscope (SEM) The Scanning Electron Microscope (SEM) is an important technique that uses a focused ion beam to produce an image of the surface of a sample. It consists essentially of an electron gun that emits electrons to the sample, condenser lenses to control the path of these electrons and detectors that measure the different signals that come from the interaction between the electrons and the sample such as secondary electrons, backscattered electrons and characteristic X-rays. Basically it can produce an image of the sample surface with very high resolution showing details down to 5 nm. In this work, the SEM is mainly used to identify and quantify the chemical composition by energy -dispersive x-ray spectroscopy (EDS) [26]. In the present work, the images of as-quenched samples, annealed samples, and samples after nanoindentation tests were taken by Zeiss NEON 40 device Focused Ion Beam/Scanning electron microscopy (FIB/SEM). Sample surface was analyzed with a SEM with electron beam energy of 15 keV. Since the samples are in ribbon form, for the characterization of the inner parts of the ribbons a trench of 15 m depth was dug with a Focused Ion Beam of Ga ions. To avoid the damage of the ribbon, a Pt protective coating was previously deposited on the surface. Image J software was used to analyze the images and to measure the size distribution of the grains in the partially crystalline and crystalline samples. The Energy-dispersive X-ray spectroscopy (EDS) was used for qualitatively analyzing the chemical composition. 18 2.6. X-ray Photoelectron Spectroscopy (XPS) The X-ray photoelectron spectroscopy (XPS) was performed with a SPECS system equipped with a Phoibos 150 MCD-9 detector and an Al anode XR50 source working at 150 W. For the high-resolution spectra, a scan step of 0.1 eV was applied. The sample spot analyzed had a diameter of 1 mm2 and the Casa XPS program (Casa software Ltd.,Teignmouth, United Kingdom) was used to evaluate the data. The spectra were normalized on the binding energy scale relative to the position of C 1 s peak at 284.8 eV and the intensity of the adventitious carbon signal was not taken into consideration. The shown spectra are not normalized on the intensity scale. 2.7. Mössbauer Spectroscopy Mössbauer Spectroscopy is a technique which can give us information about chemical, structural and magnetic properties of materials by studying the hyperfine interactions in the nuclei. It works based on the resonant emission and absorption of gamma-rays by an atomic nucleus without recoil [27]. For one nucleus to emit a 𝛾-ray and the second nucleus to absorb it, the atoms containing the two nuclei must be bonded chemically in solid state. With the help of Mössbauer spectroscopy we can measure isomer shifts and nuclear moments (i.e spin, magnetic dipole, and electric quadrupole moments). Mössbauer spectrometry can be performed with different nuclei like 57Fe,119Sn, 151Eu, 121 Sb and 161Dy but the nuclei of 57Fe is the most suitable one for achieving recoilless resonance. A basic Mössbauer spectrometer includes three important parts: the radioactive source, the absorber (the material to be studied) and a detector. If the emitting nuclei and absorbing nuclei are not in the same chemical state there would be no resonance, thus the energy of the emitted photon should be changed and this is done by Doppler effect moving the source (or the absorber). Here the radioactive source of 57Co in a Rh matrix produce the γ radiation with an energy of 14.413 keV by the decay of 57Co to excited 57Fe. The energy levels in the Fe nucleus of the sample (absorber) can change due to the presence of magnetic fields, electric field gradients or other factors, thus the energies at which the resonance takes places yield information on the so-called hyperfine parameters that are described in the following. a) Isomer shift: is a shifting in the energy levels involved in the nuclear transition due to a difference in the radius of the atom between the excited and the ground state. This parameter 19 gives information about the chemical bonding and the crystal chemical environment. b) Quadrupole splitting: if the electrical charge distribution around the Fe nucleus is not symmetric, and electric field gradient appears causing the splitting of the energy levels of the nucleus. Therefore, the quadrupole splitting can give information on the distortion of lattice surrounding the absorbent Fe nucleus. c) Magnetic splitting: the interaction between the magnetic moment of the nucleus and the magnetic moment of surrounding atoms or an externally applied field causes the separation of the nuclear spin levels, the Zeeman effect. For 57Fe, this splitting gives rise to six possible transitions. This parameter gives information on the magnetism of the sample. In this work, the Mössbauer spectra were taken in transmission mode at room temperature and pressure using a conventional constant acceleration spectrometer with a 25 mCi source of 57Co in a Rh matrix. The experimental spectra were fitted with Brand’s Normos program [28]. 2.8. Nanoindentation Nanoindentation tests are useful for evaluating various mechanical properties such as hardness (𝐻), reduced elastic modulus (𝐸𝑟), plastic energy (𝑈𝑃), elastic energy (𝑈𝐸), wearresistance and elastic recovery. Modern nanoindentation devices (see Figure 2.5) have an automatic process and are very accurate. Most of the nanoindenters use the Oliver and Pharr method [29] to analyze the force-displacement curves obtained by indentation. Figure 2.5 Nanoindentation device in Universitat Autònoma de Barcelona with a Berkovich indenter. 20 In a nanoindentation process the load-displacement curve during loading and unloading is obtained and used to calculate the hardness 𝐻 and the reduced elastic modulus 𝐸𝑟. The hardness 𝐻 of the specimen is calculated according to 𝐻=𝑃𝑚𝑎𝑥 𝐴𝐶 (2.2) where 𝑃𝑚𝑎𝑥, is the maximum indentation load and 𝐴𝑐 is the contact area under load. For a perfect Berkovich indenter, like the one used in this work, the contact area can be obtained as a function of the contact depth, ℎ𝑐, with the following equation 𝐴𝑐=𝜋.𝑡𝑎𝑛2𝜓.ℎ𝑐2 =24.56.ℎ𝑐2 (2.3) where 𝜓 = 70.32o and represents the effective semi-angle of the conical indenter equivalent to the Berkovich one. This is important because in some cases there are phenomena such as pileup and sink-in that can lead to errors in the estimated contact area. In this study we use continuous stiffness measurement (CSM), which is shown schematically in Figure 2.6, and from the load-displacement curve the slope of the unloading curve is used to calculate the stiffness values 𝑆. Figure 2.6 Continuous stiffness measurement (CSM) showing: (a) the detail of the loading curve and (b) the calculation of the stiffness by the slope of unloading curve [30]. The stiffness value is calculated by the following equation 𝑆=𝑑𝑝 𝑑ℎ=𝛽2 √𝜋𝐸𝑟.√𝐴 (2.4) 21 where the stiffness is calculated by the first derivative of the load along the load-displacement curve during the unloading process and 𝛽 depends on the geometry of the indenter (for a Berkovich indenter 𝛽 is 1.034). 𝐸𝑟 is the reduced elastic modulus that can be computed from equation 1 𝐸𝑟=1−𝜈2 𝐸+1−𝜈𝑖2 𝐸𝑖 (2.5) where v and E are the Poisson ratio and the elastic modulus of the sample and vi and 𝐸i are the same properties for the material of the indenter tip. The wear resistance of the samples can be calculated by the 𝐻/𝐸𝑟 ratio and the elastic recovery can be evaluated as the ratio between the elastic and the total (elastic + plastic) energy during the nanoindentation, 𝑈𝑒𝑙/𝑈𝑡𝑜𝑡. These energies are calculated in a nanoindentation experiment from the area between the unloading curve and the x-axis for the case of 𝑈𝑒𝑙 and from the area between the loading curve and the xaxis for the case of 𝑈𝑡𝑜𝑡 [31]. In this work, the nanoindentation measurement of the as-quenched and annealed samples was obtained employing a NHT2-Anton Paar nanoindenter on the middle of the cross-section area of the samples which were mounted on a resin base and mechanically polished with different sandpapers in a Struers Roto Pol-11 Surface Polisher (Cleveland, OH). In the final step, we used Diamond compound and 1 m sandpaper. This nanoindenter used a Berkovich pyramidalshaped diamond tip under load control mode, the maximum applied load was 10 mN with a loading segment of 30 s followed by a load holding segment of 5 s and by an unloading segment of 30 s with a thermal drift below 0.005 nm s-1. It used continuous stiffness measurement to collect material properties information as a function of displacement. 2.9. Corrosion Corrosion is an electrochemical process in which electrons are transferred between the material (mostly metal) and the environment, the corrosion reaction is basically divided in to a cathodic and an anodic reaction. In the anodic reaction, the material (metal) is oxidized and release electrons 𝑀→𝑀𝑛++𝑛𝑒− In the cathodic reaction, the reduction proceeds as follows 22 1. 𝛰2+2𝐻2𝑂+4𝑒−→4𝑂𝐻− Neutral and alkaline solutions 2. 2𝐻++2𝑒−→𝐻2↑ Hydrogen evolution (in acidic conditions) For measuring the corrosion resistance of various materials, we usually need an electrochemical cell containing a metal electrode and an electrolyte. In an electrochemical cell, a continuous electrochemical reaction take place, so chemical energy is converted in to electrochemical energy, therefore the change in energy can be calculated using 𝛥𝐺 =  𝑛𝐹𝐸 (2.6) where 𝑛 is the number of electrons involved, 𝐹 is the Faraday constant (9.648×104 C mol-1) and 𝐸 is the cell potential which is the potential difference between the two half cells. The negative sign in this equation indicates that the cell serves as an energy source. Here in the first step all the electrodes in the solution are connected together and remain for some time without being connected to an external electrical connection. By this way we can measure the potential in the circuit when it is stable, which is called the open circuit potential 𝐸𝑜𝑐. There are several ways of measuring the electrochemical properties of a sample, being two of the most important the linear polarization resistance and the electrochemical impedance spectroscopy. Linear polarization resistance. To measure the linear polarization resistance (LPR) we apply a potential above and below the open circuit potential 𝐸𝑜𝑐 and then record the current using a potentiostat. The resulting curve is called a Tafel plot, an example shown in Figure 2.7 Figure 2.7 Schematic corrosion process illustrating the anodic and cathodic current components [32] 23 In this work the potential is always measured with respect to the Ag/AgCl (3.5M in KCl) electrode that act as a reference potential. The measurement starts with a potential scan from below the open circuit potential; the potential at which the current becomes zero is called the corrosion potential 𝐸𝑐𝑜𝑟𝑟, signaling the point at which the rate of oxidation is equal to the rate of reduction. For calculating the 𝐼𝑐𝑜𝑟𝑟 we can use the Tafel slope of the anodic reaction 𝛽𝑎, which is the slope of the linear region in the anodic branch, and the Tafel slope of the cathodic reaction 𝛽𝑐 which is the slope of the linear region in the cathodic branch. The intersection of these two slopes give us the 𝐼𝑐𝑜𝑟𝑟 (see Figure 2.7) Electrochemical Impedance Spectroscopy. In an electrochemical impedance spectroscopy (EIS) experiment an alternating potential or current is applied to the sample and the response of the electrochemical system is measured. There are two ways of showing the data, the Nyquist plot or the Bode plot. In the Nyquist plot, for each excitation frequency, the real part of the impedance is plotted on the x-axis and the imaginary part is plotted on the y-axis (Figure 2.8(a)), whereas in the Bode plot the data is plotted with the logarithm of the frequency on the x-axis and the logarithm of the absolute value of the impedance (|𝑍|) and the phase-shift (𝜃) on the y-axis (Figure 2.8(b)). To simplify and better understand the EIS data, we can use the equivalent circuit diagram which contains various components such as resistors, capacitors, constant phase elements and diffusion elements that are able to model the processes taking place during the electrochemical reaction. Each of these elements are briefly explained in the following: a) b) Figure 2.8 (a) Nyquist plot (b) Bode plot [33]. 24 Resistors. We can use resistors in the equivalent circuit for the condition of electron transfer from the electrode surface to the solution or vice versa, or even to describe the movement of charges through solids and liquid phases. The impedance of a resistor is simply equal to its resistance in units of ohms (Ω), 𝑍𝑅=𝑅, and is shown in the following figure 𝑍𝑅=𝑅 (2.7) Capacitors. When non-Faradic charges accumulate at the solid/solid and solid/liquid interfaces, we can use the capacitance model. These interfaces contain contacts between adjacent metal oxide particles or nanocrystal particles, the contacts between metal oxide and a conductive substrate or the double-layer formed at the solid/solution interface. We use the following equation (2.8) and the figure in the circuit. In this equation 𝑍𝑐 is the impedance of a capacitor. 𝑍𝐶=1 𝑗𝜔𝐶= ‐𝑗 𝜔𝐶 (2.8) Constant Phase Elements (CPEs). In the electrochemical cell, which has inhomogeneities on the surface of the metal oxide electrode, such surface can be considered as a non-ideal capacitance in a double-layer at the solid/electrolyte interface. This model is mostly used for a non-ideal behavior that may be caused by surface roughness, irregularities or porosity, and it is presented by the following equation (2.9) and the figure in the circuit. Here 𝑄 is a non-ideal capacitance and has units of F and 𝛽 is an ideality factor ranging from 0 to 1. 𝑍𝐶𝑃𝐸=1 (𝑗𝜔)𝛽𝑄 (2.9) Warburg Diffusion. An important factor in EIS data analysis is the diffusion of mobile charges through metal oxide electrodes and in solution, this diffusion impedance (𝑍𝑤) is called Warburg diffusion and it is described by the following equation 𝑍𝑤=𝜎√2 (𝜔)12 ⁄ (2.10) where the 𝜎 term describes the resistance associated with diffusion as function of the concentration of charge carriers and their diffusion coefficients. 25 2.9.2. Experimental details In this thesis, the effects of B and Si on the corrosion resistance of three selected samples of HEA (A) and HEMG (AB20 and AB10Si10) are evaluated. In section 2.1. it was explained that samples were prepared by a rapid solidification technique from high purity materials in ribbon shape, thus, they have a uniform homogeneous surface with a width of 1-2 mm and thickness of 20-30 µm. The corrosion tests for all the samples were performed on the air side of the ribbons (the side of the ribbon which is in contact with air during the process of solidification in the melt spinner) which acted as the working electrode in the electrochemical cell. For preparing the ribbons for the measurements, the ribbon was connected to a pure copper (99.9%) strip by a PTFE (Polytetrafluoroethylene) tape, which has a high temperature and corrosion resistance and good stability, to get an electrical connection. The copper strip and the wheel side of the ribbon were covered completely by the PTFE tape (Figure 2.9a) in order to assure that only the sample to study is in contact with the solution. This contact was measured with ImageJ software to be able to compute the corrosion intensity density. The electrochemical tests were performed in a three-electrode corrosion cell (figure 2.9b) and carried out in 3 wt% (0.51M) NaCl solution. The reference electrode was a Ag/AgCl (3.5M in KCl) (+0.205 V vs. saturated hydrogen electrode at 25 oC) and a spiral platinum wire (with an area of ⁓3.6 cm2) was used as a counter electrode. The surface of the specimens exposed to the solution was around 0.08~1.31 cm2. The measurements were performed with a SP-200 Bioa) b) Figure 2.9 a) Preparation method of ribbon samples for corrosion testing. b) Sketch of the electrochemical cell with three electrodes. Copper strip Metallic ribbon air side PTFE tape 32 Figure 3.4 𝛥𝐻𝑚𝑖𝑥 versus 𝛿 map for as-cast high entropy alloys showing the regions favorable for obtaining crystalline solid solutions (upper left, blue dashed region) or amorphous alloys (lower right, red dashed region). The results confirm the general criteria in the case of the FeCoCrNi alloy, as reported previously [4]. The alloys with B and/or Si that show crystalline FCC or BCC solid solutions in their as-quenched states fall in the amorphous region although close to the border with the solid solution region. The multiphase compositions do not follow any clear trend as they appear mostly in the amorphous region but also outside the expected ranges for either amorphous, intermetallic or solid solution formation. Therefore, the general criteria to distinguish between the tendency to form HEAs and HEMGs is only approximately valid in these compositions, probably due to the presence of large contents of metalloid atoms. Other factors, in addition to the values of 𝛥𝐻𝑚𝑖𝑥 and 𝛿, clearly play a role in determining the GFA in this compositional system. 3.4. Thermal characterization The thermal stability of the ribbons was investigated by DSC in a continuous heating measurement up to 1500 K. The analysis of the DSC curves (Figure 3.5) shows a different crystallization behavior depending on the B and Si contents. AB20 shows a fast crystallization process and a second exothermic reaction at much higher temperatures. AB15Si5 and AB10Si10 show a crystallization reaction split in two peaks, while AB10Si15 shows a single 33 crystallization event at higher temperatures than for lower Si contents. The broad and small exothermic event for the AB10Si20 composition is a consequence of the existence of already formed nanocrystalline particles in the as-quenched ribbons as detected by XRD. Besides the number of crystallization events, the presence of Si in the samples also changes the temperature span between the first and second crystallizations, being maximum in the Si free composition and being the two crystallizations almost overlapped in the AB15Si5 and AB10Si10 samples. Figure 3.5 DSC of the as quenched samples applying a heating rate of 20 K min-1. The glass transition (𝑇𝑔) and onset of the first crystallization event (𝑇𝑥) temperatures are determined from DSC curves and from these parameters also the GFA of these HEMGs can be assessed from the supercooled liquid region (𝛥𝑇𝑥=𝑇𝑥−𝑇𝑔) and the reduced glass transition temperature (𝑇𝑟𝑔=𝑇𝑔/𝑇𝑚). The values obtained from the DSC curves are shown in Table 3.2. Alloy Label 𝜼 Tg (K) Tx1 (K) Tm (K) ΔTx (K) Trg FeCoCrNi A - - - - - - (FeCoCrNi)80B20 AB20 0 744 788 1353 44 0.55 (FeCoCrNi)80B15Si5 AB15Si5 1/3 720 769 1337 41 0.54 (FeCoCrNi)80B10Si10 AB10Si10 1 753 810 1349 57 0.56 (FeCoCrNi)75B10Si15 AB10Si15 3/2 763 871 1203 108 0.63 (FeCoCrNi)70B10Si20 AB10Si20 2 812 852 1278 - - Table 3.2. The glass transition temperature (𝑻𝒈), onset of the first crystallization event (𝑻𝒙𝟏), melting temperature (𝑻𝒎), supercooled liquid region (ΔTx) and reduced glass transition temperature (𝑻𝒓𝒈) for the completely amorphous samples, indicating also the alloy label and the Si/B ratio (𝜼). 34 The values of 𝑇𝑟𝑔 are in the range between 0.5 and 0.6. These relatively low values indicate that these compositions are not likely to be produced as bulk metallic glasses. One exception is the AB10Si15 sample that presents a slightly higher value. This sample is also the one that shows a higher thermal stability with the maximum value of 𝛥𝑇𝑥. These results suggest that there might exist a critical amount of Si that maximizes the GFA and thermal stability. 3.5. Structural characterization and effect of annealing on the asquenched ribbons In order to study the evolution of the amorphous and crystalline phases during annealing, a series of heat treatments were applied to the as-quenched samples. These treatments were chosen in accordance with the DSC curves in order to shed light on the microstructural evolution and characterize the structure at different stages of their transformation: 1) Before the glass transition temperature; 2) At the beginning and at the end of the first crystallization event; 3) At the end of the second crystallization event. Therefore, the amorphous ribbons were annealed in an Ar atmosphere with a heating rate of 6 K min-1 at different temperatures: 673, 723, 773, 823, 873, 923, 973 K. Figure 3.6 shows the XRD diffractograms for the AB20 alloy annealed at the temperatures between 673 and 973 K. Annealing at 673 K, below the glass transition temperature, does not induce any crystallization and the diffractogram still shows the broad halo characteristic of the amorphous structure. However, the diffractograms at 723 and 773 K show the appearance of broad crystalline peaks, i.e. the formation of nanometer-sized crystals, coexisting with an amorphous structure. This crystallization corresponds to the first process with onset at 𝑇𝑥1 (Figure 3.5). The presence of the crystalline phases is detected by XRD in samples annealed at temperatures below the 𝑇𝑥1 value determined by DSC at 20 K min-1 (Table 3.2); this is because of the temperature shift, expected in glass crystallization, due to the different heating rates as the annealing protocols were performed at 6 K min-1. These nanocrystalline peaks can be identified as a M3B phase, where M stands for metal and could be any combination of Fe, Co or Ni as the phases Fe3B, Ni3B and Co3B show compatible peaks with the observed ones. The JCPDS cards used for the identification are detailed in Table 3.3. It is worth to mention that the Ni and Co borides present an orthorhombic structure while the iron boride is tetragonal, showing the complexity of the crystalline phase that leads us to identify it as M3B. Although the first crystallization products in many Fe-based metallic glasses 35 are typically BCC-Fe or the Fe23B6 phase, it is not unusual to find the precipitation of Fe3B in some Fe-Si-B metallic glasses [38][39]. Phase JCPDS card Fe2B 00-036-1332 Co2B 01-075-1063 Ni2B 01-082-1697 Fe3B 00-039-1315 Co3B 00-012-0443 Ni3B 01-082-1699 Ni2Si 01-073-2093 Table 3.3. JCPDS cards used for the identification of the crystalline peaks. In the diffractograms of the samples annealed at 773 and 823 K an FCC phase begins to grow. Therefore, as a consequence of the first crystallization event, M3B and FCC phase appear in the AB alloy. The M3B phase is not stable as it disappears before the second crystallization, as can be seen in the diffractogram above 873 K. At these higher temperatures the main peak corresponds to the FCC structure and the smaller peaks can be clearly identified as M2B. In a similar way than the M3B phase, this boride can be identified with a Co2B, a Fe2B and a Ni2B phase, all with tetragonal structure. Thus, the second crystallization event induces the formation of a M2B phase while the FCC phase continues to grow. 36 Figure 3.6 XRD patterns of (FeCoCrNi)80B20 alloy after annealing at different temperatures. The crystalline peaks are identified and correspond to an FCC phase and two borides. The XRD diffractogram of AB10Si10 alloy annealed at the same temperatures is shown in figure 3.7. Below 773 K, the diffractogram shows completely amorphous samples and at 773 K some nanocrystals begin to form that can be assigned to a BCC phase together with a peak that can be ascribed to a M2(B,Si) phase, that remains visible at higher annealing temperatures including the fully crystallized sample. In this case, the Si can be incorporated in the structure of the boride but can also be present in a Ni2Si phase. Annealing at 823 K yields the formation of an FCC phase together with a new boride, the M2(B,Si) phase that continues to grow at higher temperatures. At this temperature the peaks corresponding to the BCC phase are still present. Further annealing induces the disappearance of the BCC phase and the growth of the FCC and M2(B,Si) phases. Figure 3.7 XRD patterns of (FeCoCrNi)80B10Si10 alloy after annealing at different temperatures. The crystalline peaks are identified and correspond to an FCC, BCC and M2(B,Si) phase. 3.6. In-situ characterization of the crystallization In the previous sections we presented the XRD patterns for the AB20 and AB10Si10 in the as-quenched state and annealed at different temperatures. The conventional laboratory XRD 37 technique is useful to give information about the phases that appear during each crystallization but give little information about the microstructural change with time. Thus, the structural properties of the ribbons were investigated by Synchrotron XRD at the MSPD beamline of ALBA Synchrotron. The X-ray synchrotron experiments were performed to measure in situ the crystallization process. The ribbon was irradiated with monochromatic radiation of 𝜆= 0.4246 Å during a heating up from room temperature to 793 K at 20 K min-1 and held at this temperature for 2h and cooled down to room temperature at 100 K min-1. Patterns were collected with a time acquisition of 1 s every 5 s using a Rayonix CCD detector. A diffraction without sample was taken at room temperature to have the background signal. The diffraction pattern at each temperature and the one of the backgrounds has been subtracted after intensity normalization. The diffraction was performed on the as-quenched ribbons; this may introduce some artefacts in the detected intensities of the Bragg reflections due to possible texture of the growing crystalline phases. However, the production of powder for the XRD analysis was discarded in order to avoid structural changes originated due to the milling process. The crystallization of the samples was analyzed in order to assess the role of Si and B on the crystallization path and characterize the emerging phases. According to the value of the ratio 𝜂 = Si/B, and the DSC signal (Figure 3.5) we can distinguish two different regions. For 𝜂 ≤ 1, there are two crystallization events before 873 K, except for 𝜂 = 0, where the second crystallization occurs above 873 K. For values of 𝜂 larger than 1, there is only one crystallization event and, for 𝜂 = 2 the crystals are already present in the as-quenched sample. Figure 3.8 shows the crystallization for the case 𝜂 = 0. The left panel shows the evolution of the intensity of the diffraction peaks with the temperature. The amorphous character of the as-quenched state is clearly seen by the low intensity of the broad peak between 10o and 15o. This broad peak is replaced by sharp lines at temperatures close to 790 K signaling the nucleation and growth of the crystalline phases. The continuity of the lines indicates the existence of only one crystallization event while the increase in brightness of the lines shows the increase in volume fraction of the crystalline phases as the holding time at the maximum temperature increases. Therefore, with a holding time of 120 minutes at 793 K the second crystallization event detected by DSC is not reached in the X-ray diffractograms. The right panel of Fig. 3.8 shows the intensity versus the diffraction angle for selected temperatures from the initial amorphous state up to the final crystallized state. 38 Figure 3.8 Crystallization of the AB20 sample (𝜂 = 0). Left panel: 2D map of the crystalline peaks showing their intensity (in a color scale) versus the diffraction angle 2𝜃 (x axis) and the temperature 𝑇 (y axis). Right panel: diffractograms of selected temperatures, from the initial amorphous phase to the highest temperature with phase identification. The crystallization pattern is complex with multiple low intensity peaks. The main phases that can be identified are a solid solution with an FCC structure, the observed peaks correspond to (111), (200), (220) and (311) reflections, and two borides: Fe3B and CoB. However, the intrinsic chemical disorder characteristic of high-entropy alloys tends to produce crystalline phases with a mixture of the main elements, thus it can be expected that the Fe3B and CoB phase will contain other metal elements. This complexity of the crystalline phases is analyzed in detail by Mössbauer spectroscopy in the next section, showing that the Mössbauer spectra of the boride phases presented hyperfine parameters that differed from the ones corresponding to a pure Fe-boride phase, thus confirming the presence of other elements (Cr, Co or Ni) in the structure. The overall structure of the phases can be identified but, similarly to other works in this type of materials [38][39], the exact distribution of metal atoms inside the borides (or silicides, in the case of Si containing samples below) cannot be determined by the techniques used in this study. Besides the two identified borides, the three more intense peaks between 18 and 20o can be attributed to a more complex boride close to the Fe3Ni3B structure and to the intermetallic phase Cr3Ni2 although the identification of this latter phase is ambiguous. As explained before, in this experiment the second crystallization event observed by DSC cannot be reached, although in the previous section it was shown that increasing the temperature induces the formation of an M2B phase and the disappearance of the M3B phase. Thus, the first crystallization can be summarized in the following way: 𝜂 = 0: Amorphous → FCC + MB + M3B + M6B, 39 where M stands for metallic atoms. The crystallization for the case 𝜂 = 1/3 can be seen in figure 3.9. The 2D map clearly shows that two crystallizations processes take place in this composition. Firstly, the amorphous phase crystallizes in a BCC structure, likely a solid solution, with the peaks at 12o, 17o and 21o corresponding to the reflections (110), (200) and (211), respectively. After reaching the maximum temperature in the furnace, the second crystallization takes place with the nucleation and growth of an FCC structure together with two borides, CrFeB and FeB. There is no clear set of peaks that can be ascribed to a silicide phase, thus it is reasonable to assume that the small amount of Si atoms is present in the other crystalline phases or in a residual amorphous fraction. However, the small shoulder at around 11o together with the two more intense peaks could also be compatible with an Fe4.9Si2B phase. The crystallization path for this case is, thus: 𝜂 = 1/3: Amorphous → BCC → BCC + FCC + MB(Si) + M2B(Si), where (Si) means the possible presence of this element in the structure, although not confirmed. Figure 3.9 Crystallization of the AB15Si5 sample (𝜂 = 1/3). Left panel: 2D map of the crystalline peaks showing their intensity (in a color scale) versus the diffraction angle 2𝜃 (x axis) and the temperature 𝑇 (y axis). Right panel: diffractograms of selected temperatures from the initial amorphous phase to the highest temperature with phase identification. The composition with 𝜂 = 1 also presents a double crystallization reaction (see Figure 3.10). The first one takes places below 793 K and, like the previous composition, it can be ascribed to a solid solution with a BCC structure. However, contrary to 𝜂 =1/3, in the case of 𝜂 =1 there is a simultaneous crystallization of boride and silicide phases. In particular, peaks corresponding to Fe2B, Ni2Si and Cr1.64Fe0.35B0.96 can be globally identified as phases with an M2(B,Si) structure, although some of the peaks can also be ascribed to a Co2B phase. The second crystallization event maintains the previous phases and it induces the nucleation and 40 growth of a solid solution with an FCC structure, resulting in the following global crystallization path: 𝜂 = 1: Amorphous → BCC + M2(B,Si) →BCC + FCC + M2(B,Si) Figure3.10 Crystallization of the AB10Si10 sample (𝜂 = 1). Left panel: 2D map of the crystalline peaks showing their intensity (in a color scale) versus the diffraction angle 2𝜃 (x axis) and the temperature 𝑇 (y axis). Right panel: diffractograms of selected temperatures from the initial amorphous phase to the highest temperature with phase identification. The last of the studied compositions with a completely amorphous structure in the as-quenched state corresponds to the case 𝜂 = 3/2 and, contrary to the previous cases, it only presents one crystallization event in Figure 3.4. From the 2D map of figure 3.11 it could be seen that there are two different crystallization stages, however, looking at the selected diffractograms in the right-side panel it can be seen that all the peaks are already present at the beginning of the crystallization, although some of them are masked by the halo corresponding to the amorphous matrix in which the crystalline phases grow. In a similar way than in the case 𝜂 = 1, the main existing crystalline phase is a solid solution with a BCC structure that grows simultaneously with borides and silicides. In particular, the peaks can be identified with Ni2Si and Cr2B phases. The peaks at around 11o and 17o are not identified but can be compatible with Fe5Si3 or Ni3Si phases if some texture is assumed. Unlike the other compositions with less Si, in this case there is no growth of an FCC structure. What it can be observed, however, is that after a relatively short time at 793 K (approximately 30 minutes) the peaks corresponding to the BCC solid solution begin to decrease in intensity at the expenses of borides and silicides. Therefore, in this composition the crystallization path is more simple and can be summarized as follows: 𝜂 = 3/2: Amorphous→ BCC + M2(Si,B) 41 Figure3.11 Crystallization of the AB10Si15 sample (𝜂 = 3/2). Left panel: 2D map of the crystalline peaks showing their intensity (in a color scale) versus the diffraction angle 2𝜃 (x axis) and the temperature 𝑇 (y axis). Right panel: diffractograms of selected times after reaching the maximum temperature in the hot stage. The last of the studied compositions corresponds to the case 𝜂 = 2, which is partially crystalline in the as-quenched state, although the majority of the as-quenched structure remains amorphous (see Figure 3.12). Despite this double character of the structure, the study of the crystalline phases is interesting in order to have a better picture of the effect of Si in the crystallization path of this family of HEMG. The crystals present in the as-quenched state can be identified as a solid solution with a BCC structure that continuously grows when increasing temperature. After the full treatment of the sample (120 minutes at 793 K) besides the BCC phase, several peaks appear that are compatible with a Ni2Si phase. There are no clear peaks corresponding to borides, although the peaks at 11.4o and 12.2o are consistent with a Ni2B phase. Thus, the crystallization path is: 𝜂 = 2: Amorphous + BCC → BCC + M2(Si, B). 48 ABSi amorphous  BCC + M2(B,Si)  FCC + BCC + M2(B,Si)  FCC + M2(B,Si) 3.8. Effect of annealing on the particle size distribution The annealing temperature is shown to have a significant influence on the particle size distribution and morphology of the ribbons. The SEM images of the interior of the AB20 and AB10Si10 ribbons at room temperature and after annealing at different temperature up to 973K are shown in figure 3.18. The first two figures on top show the AB20 and AB10Si10 asprepared ribbons at room temperature with a smooth and homogeneous cross section characteristic of the amorphous materials. Although, as detected by Mössbauer, the crystallization process begins at 723 K, the formed crystals are too small to be noticed in the SEM images. In samples annealed at 823 K a clear crystalline structure, uniformly distributed, is observed with a typical grain size that continuously grows from 923 K to 973 K. Comparing the last rows of Figure 3.16 it can be concluded that the presence of Si hinders the growth of the grains resulting in a sample with a smaller crystals compared to the AB sample. AB20 AB10 as-q 823K 49 873 K 973 K Figure 3.18 SEM images of the HEMGs alloys (FeCoCrNi)80B20 (left column) and (FeCoCrNi)80B10Si10 (right column) in their as-quenched state (first row) and after annealing at 823, 873 and 973 K. To further study the effect of Si in the microstructure, the grain size distribution of both alloys as a function of the annealing temperature was computed from the SEM images of Figure 3.18. The average grain size was calculated by ImageJ software using the following procedure: i) setting the scale according to scale of the photo; ii) measuring each grain 5 times from different directions and iii) taking average of the length for each grain. The AB alloy (Figure 3.19 top) grows from ~69 nm at 823 K to ~185 nm at 973 K. In all the temperatures the width of the distribution is large, especially after 823 K, indicating a microstructure without a welldefined characteristic length. In contrast, Figure 3.19 bottom shows the grain size distribution for the ABSi alloy, with a lower average grain size that grows from ~37 nm at 823 K to ~135 nm at 973 K and with a narrower width below 973 K. Therefore, the presence of Si instead of B in the microstructure not only reduces almost to a half the average size of the crystals but also produces a more uniform microstructure. 50 Figure 3.19 Distribution of grain sizes in (FeCoCrNi)80B20 (top) and (FeCoCrNi)80B10Si10 (bottom) after annealing at different temperatures. 3.9. Discussion In this chapter, we have presented two different alloys and in each section their properties and the effect of the annealing and the elemental composition on them has been discussed. In the first step, (FeCoCrNi)80B20 and (FeCoCrNi)80B10Si10 were produced and it has been shown that both of them followed similar thermal behavior but the partial substitution of B with Si causes the increase of both the glass transition temperature and the temperature of the first crystallization event while it reduces the temperature of the second crystallization. Moreover, it shows a slight increase in the supercooled liquid region (𝛥𝑇𝑥) after adding Si, therefore this element increases the thermal stability of the alloy. Regarding the structure of these alloys at room temperature, it has been shown that both, AB and ABSi were amorphous at room temperature. By annealing the AB sample, the FCC and M3B phases appeared as a first 51 crystallization product at 823K whereas M3B was not a stable phase and it disappeared before the second crystallization. After the second crystallization, the FCC phase still grows beside the formation of M2B and this M could correspond to the different metals (Fe,Co,Ni). On the other side, for the ABSi alloy, the x-ray diffraction pattern showed a completely amorphous phase below 773K and by increasing the temperature the nucleation and growth of some nanocrystals like BCC and M2(B,Si) begin. By increasing the temperature to above 823K the formation of FCC and some new boride phase started. At higher temperature, the BCC phase is still there and the growth of FCC and M2(B,Si) continued. An addition of 10 at% of B and 15 at% of Si to the base composition (equiatomic FeCoCrNi) maximize the glass forming ability and thermal stability of the amorphous phase. Also we studied the crystallization path of other compositions in this new family of high entropy metallic glasses (AB20, AB10Si10, AB10Si20, AB10Si15 and AB15Si5) characterizing their structure by in-situ X-ray diffraction at ALBA synchrotron and we reach the following conclusions. For Si/B ratios lower or equal than one the alloys present two crystallization events while for higher values of this ratio there is only one crystallization peak. Moreover, as the amount of Si increases the main crystalline phase changes from FCC to BCC. In the last section of the chapter we discussed the microstructure and the effect of the heat treatment on it. From a microstructural point of view there is a change in the size of nanocrystalline particles in both alloys. According to our study, in the AB alloy the average grain size at 823K was ~ 69 nm and by increasing the annealing temperature it will reach to ~185 nm at 973 K while in the ABSi alloy, the average grain size grows from ~ 37 nm at 823K to ~135 nm at 973K. Eventually, the replacement of part of B with Si causes the microstructure to reduce the size as well as to produce a more uniform microstructure. 52 53 CHAPTER 4. MECHANICAL CHARACTERIZATION The aim of this chapter is to study the mechanical properties and the deformation process of the produced alloys by means of nanoindentation. This technique offers many advantages and it requires only a small volume of test material [40]. Several mechanical properties related to the elastic and plastic deformation response can be assessed from the load-displacement (𝑃-ℎ) curve recorded during the nanoindentation process [41] that allow us to determine the hardness (𝐻), the reduced Young’s modulus (𝐸𝑟), the plastic energy (𝑈𝑝), the elastic energy (𝑈𝑒𝑙 ), the wear resistance (𝐻/𝐸𝑟 ) and the elastic recovery (𝑈𝑒𝑙/𝑈𝑡𝑜𝑡). In this chapter, of all the produced compositions we selected the ones with best glass-forming ability and the base alloy, the latter to serve us as a reference. These compositions are: A, AB20, AB10Si10, AB10Si20, AB15Si5 and AB10Si15, emphasized with a circle in Figure 3.3 of the previous chapter. 4.1. Nanoindentation The mechanical properties of the as quenched and the annealed samples are obtained by the nanoindentation equipment described in section 2.7. Figure 4.1 (top) shows the loaddisplacement curves of all the samples in the as-quenched state, including composition A for reference. In addition to the different penetration depths and corresponding hardnesses, the slope of the unloading part of the indentation curves, related to the contact stiffness, indicates a different value of the reduced Young’s modulus. From these curves the hardness and 𝐸𝑟 can be calculated and the values are shown in Figure 4.1 (bottom) and Table 4.1. Figure 4.1 (bottom) shows the change of these magnitudes as a function of the composition for all the as-quenched samples. The large increase in hardness from the base composition (⁓ 4 GPa) to the B and Si containing alloys (⁓14-18 GPa) can be explained by the strong nature of the (B,Si)-M interatomic bonds and the change from an FCC crystalline to an amorphous structure, where the absence of long-range order and crystalline defects increases the resistance to plastic deformation. There is no clear trend in the values of 𝐻 or 𝐸𝑟 as a function of composition, all amorphous samples showing outstandingly high hardness values, three times as hard as stainless steels [42]. 54 Figure 4.1 Load-displacement nanoindentation curves (top) and variation of hardness (𝐻) and reduced Young’s modulus (𝐸𝑟) (bottom) for as quenched ribbons at room temperature. However, there is a clear increase of these magnitudes for AB10Si20, i.e. the alloy that has the major content of B and Si and, in addition, contains some nanocrystals embedded in its amorphous structure as discussed in the previous section. It is expected that the elastic properties of a material are related to its atomic bonding in a way that compositions with stronger bonds exhibit higher values of 𝐸𝑟 [43]. There are two possible contributions to the maximum values of 𝐻 and 𝐸𝑟 in the AB10Si20 alloy. On the one hand, the large amount of Si creates stronger covalent bonds between this element and the metallic elements. On the other hand, the interplay between the amorphous matrix and the nanocrystalline inclusions can enhance the mechanical properties [35][44]. 55 Samples 𝑯 (GPa) 𝑬𝒓 (GPa) 𝑼𝒑 (nJ) 𝑼𝒆𝒍 (nJ) 𝑼𝒕𝒐𝒕 (nJ) 𝑯/𝑬𝒓 𝑼𝒆𝒍 𝑼𝒕𝒐𝒕 ⁄ A 4.15 ± 0.12 121.9 ± 4.1 1.04 ± 0.029 0.247 ± 0.008 1.28 ± 0.04 0.034 ± 0.002 0.190 ± 0.011 AB20 14.40 ± 0.53 167.1 ± 4.8 0.353 ± 0.002 0.343 ± 0.002 0.690 ± 0.004 0.086 ± 0.005 0.490 ± 0.005 AB15Si5 12.50 ± 0.46 149.6 ± 6.5 0.405 ± 0.006 0.351 ± 0.002 0.750 ± 0.008 0.084 ± 0.006 0.460 ± 0.007 AB10Si10 15.50 ± 0.54 162.3 ± 1.8 0.367 ± 0.007 0.356 ± 0.003 0.72 ± 0.01 0.09 ± 0.003 0.49 ± 0.01 AB10Si15 12.6 ± 0.4 155.4 ± 2.1 0.401 ± 0.002 0.345 ± 0.001 0.740 ±0.003 0.081 ± 0.003 0.46 ± 0.007 AB10Si20 17.7 ± 0.7 176.3 ± 3.6 0.279 ± 0.020 0.361 ± 0.002 0.64 ± 0.03 0.100 ± 0.005 0.56 ± 0.025 Table 4.1. Hardness (𝐻), reduced Young’s modulus (𝐸𝑟), plastic energy (𝑈𝑝), elastic energy (𝑈𝑒𝑙), wear resistance (𝐻/𝐸𝑟) and elastic recovery (𝑈𝑒𝑙/𝑈𝑡𝑜𝑡) measured by nanoindentation. We can compare the hardness and elastic modulus values of the as-quenched alloys studied here with other high-entropy alloys of similar compositions probed also by nanoindentation [45][46]. In these references the maximum loads were between 10 and 30 mN, similar to the ones applied in the present work, although some differences could be expected due to the measures done at a strain rates of 5x10-2 s-1 in the case of ref. [45]. Figure 4.2 displays the values of the Young’s modulus versus hardness for several alloys that also contain Fe, Co and Cr in the base composition and additions of other elements like, Cu, Si or Al to tailor their properties. From this figure, it is clear that the micro alloying with Al and the incorporation of Mn in the base composition are beneficial for the elastic modulus but it produces only a moderate increase in the hardness. All these compositions present a diversity of structures and microstructures depending on the particular composition and alloying element and comprise FCC, BCC, mixture of BCC and FCC and also a mixture of FCC and HCP phases. The composition with the highest elastic modulus is the Al3(FeCoCrNiCu) that consists of a BCC phase in a dendritic configuration. Accordingly, it could be concluded that more than the particular crystalline configuration of the alloy, the most determinant factor that contributes to its mechanical behavior is the microstructural configuration of the main phase. The use of important amounts of B and/or Si significantly improves (almost doubles) the hardness, while keeping the elastic modulus almost constant. This increase of hardness opens a new route to design new high-entropy alloys with improved properties seeking for a combination of 56 constituent elements capable to amorphize the structure or induce a change in the microstructure. Figure 4.2 Comparison of Young modulus and hardness for several compositions [45][46]. The circle indicates the compositions investigated in this study. In order to study the effect of nanocrystallization on the mechanical properties, two of the compositions, AB20 and AB10Si10, were selected and nanoindentation experiments were performed at different stages of their crystallization process. These two alloys have the same ratio of metallic and metalloid elements but different crystallization routes, forming FCC or BCC crystals as described in the previous chapter. Specifically, the samples were annealed up to different temperatures to examine their properties as a function of their degree of crystallization. The set of annealing temperatures spans from below the glass transition up to full crystallization. It is well known that below 𝑇𝑔 a heat treatment causes changes in the physical and mechanical properties of glasses due to structural relaxation. The decrease of excess free volume during this process usually causes a reduction of atomic mobility an increase of Young’s modulus and hardness, and a loss of ductility of metallic glasses [43][40][47][41]. 57 Figure 4.3 Variation of hardness (𝐻) and reduced elastic modulus (𝐸𝑟) as a function of annealing temperature for AB20 (top) and AB10Si10 (bottom). The results of 𝐻 and 𝐸𝑟 for the two compositions are shown in Figure 4.3 and are detailed in Table 4.2. Figure 4.3 (top), corresponding to the Si-free material, shows a progressive increase of both 𝐻 and 𝐸𝑟, up to the sample annealed at 773 K. According to our previous characterization in chapter 3, after an annealing at 773 K the sample is composed of nanocrystals (FCC and M3B) embedded in an amorphous matrix. At this point, the maximum values of the hardness and reduced Young’s modulus are reached, being 24 GPa and 215 GPa, respectively. The increase of the mechanical properties between the as-quenched and the annealed amorphous samples can be attributed to the annihilation of excess free volume associated to the structural relaxation of the glass [47]. At higher annealing temperatures, the crystallization of the hard M3B phase also contributes to the increase in the strength of the sample [44]. At annealing temperatures higher than 773 K, 𝐸𝑟 remains almost constant while 64 65 CHAPTER5.ELECTROCHEMICAL CHARACTERIZATION Corrosion-related accidents cause major economic loses and are a huge concern for the safety of personnel and property. The direct annual cost of corrosion across the globe was reported to be approximately 3 percent of global gross domestic product (GDP) [64]. In the United States, around US$ 2-4 trilions is lost due to corrosion-related failures each decade. Therefore, it is necessary to study corrosion and its different forms to design and choose suitable materials for specific applications and improve safety standards. Corrosion is the chemical reaction of a material within the environment which causes its degradation. This reaction can be sorted in numerous ways including the nature of the corroding agent of which there are two categories, wet and dry. Wet corrosion refers to the corrosion of a material in aqueous environment with an electrolyte and dry corrosion refers to the corrosion of a material in a gaseous environment. Another way to classify corrosion is the appearance of the corroded metal (uniform or localized). Uniform corrosion take places when the corrosion reaction happens over all the surface of the material while localized corrosion occurs at specific locations [65]. Other mechanisms of corrosion can be categorized under the names of electrochemical and direct chemical. Direct chemical corrosion is an attack resulting from direct exposure of a bare surface to a caustic liquid or gaseous agent occurring simultaneously at the same point while in an electrochemical attack the anodic and cathodic changes may take place a measurable distance apart. Electrochemical is the most common corrosion process and it happens when two or more electrochemical reaction occur, including at the minimum one oxidation and one reduction reactions. In overall there are two main requirements for the reaction: a) there must be an anode and a cathode to provide an area for oxidation and reduction and b) there must be a pathway for electrons to flow from the anode to the cathode. If any of these conditions are removed, the corrosion process stops. The anodic and cathodic sites can be created on the surface of a metal due to heterogeneities (like composition and grain size differences, surface roughness, impurities or inclusions, localized stresses or dislocation arrays) or between two dissimilar metals exposed to the corrosive environment. Metals can have three different responses to an electrochemical reaction according to the environment: it can show immune, active or passive reaction. Immune reaction happens when materials are thermodynamically stable in an environment and, therefore, they will not corrode. Active 66 metals go through the corrosion reaction with the environment and form corrosion products. In a passive reaction, corrosion at first happens but the resulting product is insoluble and acts as a protective film. This coating layer dramatically reduces the rate of corrosion but if it is broken, the metal turns over to an active reaction. Corrosion can be avoided by the following methods: alloy additions, coatings and inhibitors or design modification. It was discovered that the strongest influence on the corrosion behavior is caused by the chemical composition and microstructure in the alloys. Proper material selection is one of the most effective strategies to reduce the corrosion rate. Metallic glasses, hold tremendous potential as corrosion-resistance alloys (CRAs) and a series of metallic glasses based on copper, palladium, zirconium, titanium, magnesium and iron have been successfully produced [66]. In the case of a positive influence of alloying on the corrosion properties of alloys, B and Si have been shown as good alloying elements [67],[68]. Therefore, the aim of this chapter is to assess the effects of B and Si against corrosion resistance of some selected HEAMG’s compositions produced in this thesis, in particular (FeCoCrNi)80B20 and (FeCoCrNi)80B10Si10, which were produced by rapid-solidification technique and were labelled in chapter 3 as AB20 and AB10Si10 samples, respectively. The corrosion resistance of these two samples will be compared with the one of the HEA (FeCoCrNi)100, labelled A. It is well known that all sources of localized corrosion disappear in the absence of common structural defects such as grain boundaries, dislocations and segregations [68][69],[70],[71],[72]. Moreover, previous research showed the effect of structure, chemical composition and material homogeneity on the electrochemical behavior of amorphous alloys [73]. Considering the remarkable corrosion resistance of Fe, Ni, and Cobased amorphous alloys, it is important to study the effect of metalloid elements (B, Si) as well as other alloying elements (Cr, Mo, ..) that could improve the glass forming ability or the corrosion resistance. The main studies on the subject have shown that the corrosion resistance of Fe, Cu, Ni and Co-based amorphous alloys in sulphate and chloride media is largely affected by alloying with various additional elements such as Cr, Mo, Ni [72],[73]. 5.1. Electrochemical measurements in NaCl solution The potentiodynamic polarization curves for the A, AB20, and AB10Si10 samples in 3% NaCl solution are shown in Figure5.1. The corrosion potential (Ecorr) and corrosion current density (Icorr) were determined by extrapolating the Tafel curves and are summarized in Table5 67 1. The values of Ecorr and Icorr of the as-quenched ribbons change with the amount of B and Si. Here, AB20 has the highest Ecorr and lowest Icorr. With the reduction in B and increase in Si, the Ecorr decreases while Icorr is kept constant. The crystalline ribbon without B and Si, which we take as the reference material, has the lowest Ecorr. It is well known that corrosion is more likely to occur at grain boundaries, defects, and regions where there is large segregation of elements, thus the amorphous nature of the AB20 and AB10Si10 samples promotes chemical and microstructural homogeneity and improves the corrosion resistance [74]. Moreover, the action of B as a corrosion inhibitor is also reported by several authors [75],[76]. Accordingly, the Ecorr value of the AB10Si10 is slightly lower than the one of the AB20 alloy, with the same corrosion current density. These electrochemical parameters can be compared with the ones corresponding to other high-entropy alloys. If we compare them with crystalline HEA, the results shown here present a more noble Ecorr and lower Icorr [77],[78],[79],[80] and they also show an improvement with respect to other similar high-entropy bulk metallic glasses that show Ecorr values between −23 and 77 mV [81]. Figure5.1 Potentiodynamic polarization curves of as quenched A, AB20 and AB10Si10 ribbons in 3 wt% NaCl. The inset show the evolution with time of the OCP. Alloy 𝑬𝒄𝒐𝒓𝒓 (mV) 𝑰𝒄𝒐𝒓𝒓(nA 𝐜𝐦−𝟐) A 70 0.09 AB20 170 0.06 AB10Si10 150 0.06 Table 5.1 Summary of the quantitative analysis of the potentiodynamic polarization test of the A, AB20 and AB10Si10 samples. The estimated error for the corrosion potential is ±10 mV while for the corrosion current is ±0.01 nA cm-2. 68 At this point, a question arise about which factor plays a more important role in corrosion resistance. As commented on previously, some studies explain the improvement in the corrosion resistance because of the high-entropy effect that reduces the mobility of the atoms [82]. We can compute the entropies of these alloys and distinguish between the configurational entropy and the mismatch entropy [83]; the configurational entropy is simply proportional to the ln(N), where N is the number of components, thus we have an increase in this entropy from the A alloy (N = 4) to the AB10Si10 alloy (N = 6). The mismatch entropy has been shown to be proportional to the delta parameter that is a measure of the atomic size difference between the constituent atoms, and this value increase from 0.3% for the A alloy to 14.70% for the AB20 and 10.72% for AB10Si10 [84]. Therefore, from these values and the electrochemical parameters, it would seem that the mismatch entropy is the main factor affecting the corrosion resistance, but this analysis does not consider the fact that AB20 and AB10Si10 are amorphous samples with a disordered structure. Thus, in order to clarify this point, a systematic analysis of several families of HEAs and HEMGs should be performed but this is out of the scope of this paper. 5.2. Electrochemical Impedance Spectroscopy (EIS) measurements Figure 5.2a,b shows the Nyquist and Bode plots, respectively, of the EIS measurements on the three samples. The absolute value of the impedance (|z|) at very low frequencies (ω  0) represents the polarization resistance (Rp) which is the transition resistance between the electrodes and the electrolyte while the value at very high frequencies (ω  ∞) represents the solution resistance, (Rs) [85]. As can be seen in Figure 5.2a, the three samples show similar behavior. The phase angle shown in Figure 5.2b and defined as: ϕ=tan−1(𝐼𝑚(𝑍) 𝑅𝑒(𝑍)) (5.1) presents some differences. The broadened base of the phase angle and its magnitude are a signal of the pseudo-capacitive behavior, which can be modelled by using a constant phase element [85,86]. Here the single local minimum indicates the presence of a capacitance element that can be modelled by a single capacitor or two capacitors of similar magnitude. For all of these three alloys, these capacitors can be passive films or an electrical double layer capacitance. In order to quantitatively assess the differences between the alloys, the Nyquist and Bode plots have been fitted to an equivalent circuit consisting of one resistor in series with a parallel 69 combination of a constant phase element and a Warburg element and also in series with a parallel combination of a resistor and a capacitor as illustrated in Figure 5.3b. The fitting was performed with EC-Lab v11.10 software and the obtained parameters from the fitting are provided in Table 5.2. The left part of the circuit, which contains the resistor R2 and the capacitor C2, is related to the movement of mobile charges trough the solid and liquid phases and the non-faradic charge accumulation at the solid/liquid interface of the electrode. (a) (b) Figure 5.2 (a) Nyquist plot of A, AB20 and AB10Si10 impedance fitted by the inset equivalent circuit model. Raw data are dots and the fitting result is represented by lines of the same color and (b) Bode plot of A, AB20 and AB10Si10 This charge accumulation in the interface constitutes what is commonly known as the capacitive double layer (shown in Figure 5.3) where three different regions can be distinguished: a first layer that mainly contains polar water molecules and adsorbed anions, called the inner Helmholtz plane (IHP), a second layer with fully hydrated cations, called the outer Helmholtz plane (OHP) and a final diffuse layer composed of hydrated anions and cations [87]. From the values of Table 5.2, it is evident that the amorphous alloys present higher corrosion resistance as the value of the capacitance is the lowest with respect to the crystalline one. Therefore, it can be stated that in the amorphous alloys, the formation of the double layer capacitance increases the corrosion resistance. However, the overall electrochemical behavior also depends on the inhomogeneities of the electrode surfaces. 70 Figure 5.3 (a) Schematic representation of a metal electrode surface in the solution and (b) the equivalent electrical circuit used to fit the impedance data. R1 and R2 are the electrolyte and charge transfer resistance, respectively. Q1 is the constant phase element, W1 is a Warburg element and C2 is a capacitor. Electrode surface irregularities play a significant role in the electrochemical response. There are several factors which cause these irregularities, such as surface roughness and chemical heterogeneities (that include differences in the constituent elements, chemical impurities, and surface band impurities and coatings). Moreover, solid electrodes are not smooth; they exhibit complex surface morphologies with a varying degree of irregular interfaces (i.e., rough, porous, and partially active interfaces). All these effects are modelled by the middle part of the equivalent circuit, which contains a constant phase element (Q1) that corresponds to inhomogeneities in the surface on the atomic and nanoscopic scale (roughness) and crystallographic disorder (due to anisotropic surface atomic structure) of the metal oxide electrode, and a Warburg diffusion element (W) that corresponds to the diffusion of mobile charges within the metal oxide electrode in the solution, respectively. The higher values of Q1 for the A alloy reflect the higher inhomogeneity of this alloy and its lower resistance to corrosion. Finally, the right part of the circuit, R1, is related to the resistance of the electrolyte Specifically Adsorbed Anion Metal Solution IH P OHP Diffusion region Solvent molecule Solvated Cation Electrolyte solution Electrode surface and Double layer capacitor R1 Q1 W1 R2 C2 Oxygen Evolution and Diffuse layer Q1 a) b) 71 solution that in this case is higher for the AB10Si10 alloy reflecting better electrochemical behavior. Alloy 𝑹𝟏(𝛀) 𝑹𝟐(𝐤𝛀) 𝑪𝟐 (F) 𝑸𝟏(𝐅𝐬(𝒂−𝟏)) 𝒂 𝑾𝟏(𝛀𝐬−𝟏/𝟐) A 33.45 1149 1.697×10−6 1.335×10−6 0.906 3.219×106 AB20 34.41 21310 0.854×10−6 0.991×10−6 0.913 1.844×106 AB10Si10 100.4 258800 0.156×10−6 0.200×10−6 0.927 5.543×106 Table 5.2 Parameters from the fitting of the Nyquist plot to an equivalent electrical circuit. 5.3. XPS analysis of A, AB20 and AB10Si10 alloys immersed in NaCl The corrosion resistance of HEMGs in a particular corrosion environment will be determined by how their surface reacts to the corrosion agent and by the composition of the passive film that can be formed on the surface. Therefore, X-ray photoelectron spectroscopy (XPS) is used in this work to evaluate the surface to obtain information about the effect of anode polarization on the surface composition and determine the oxidation state of the different elements [88]. The use of small amounts of Al, Cr, or Si has been shown to improve oxidation resistance in conventional alloys due to the formation of a stable oxide layer on the surface [89]. On the contrary, Fe-based HEAs do not present these restrictions on the amount of other elements and, thus, Al, Cr, or Si may be present in higher concentrations in order to facilitate the formation of the oxide films [90]. The XPS spectra of our samples are shown in Figures 5.4a,5.5a, and5.6a, for the A, AB20, and AB10Si10 samples, respectively. The figures show the survey spectrum (low-resolution measurements in a large energy range) in panel (a) and the high-resolution emission peaks associated with the Fe 2p, Cr 2p, and O 1s core electron levels in panels (b), (c), and (d), respectively. Moreover, Figures5.5 and5.6 also include in panel (e) the B 1s core electron level and the Si 2p core electron level emission peak, respectively. The different dashed lines mark the theoretical energy of different oxidation states of the corresponding element and are used as a reference. The C 1s line is due to the carbon present in the sample holder and is used for the energy calibration. The presence of O in the long-range energy spectra of the three alloys is observed before and after polarization measurements, suggesting that an oxide film is spontaneously formed on the sample surface 72 before immersion in the NaCl solution. Figure5.4a shows the normalized XPS spectrum of sample A after immersion in a 3 wt.% solution of NaCl, indicating the presence of Fe, Cr, Co, Ni, and O, although the peaks associated to Co and Ni are barely visible. High-resolution spectra at the energy of the Fe 2p peak reveals that Fe is present on the surface as Fe3+ with a binding energy of 710.15 eV (Figure5.5b). On the other hand, the Cr 2p peaks are at 573.29 and 575.91 corresponding to metallic Cr and Cr3+, respectively (Figure5.5c). The presence of Cr3+ is considered to be crucial to the quality of the passivation film for the A composition. In addition, the binding energy peaks of O 1s represents a M-O compound characteristic peak corresponding to O−2 species (529.54 eV) and a M-(O-H)n compound characteristic peak corresponding to O-H species (532.06 eV) (Figure5.4d). Figure 5.4 (a) Normalised XPS spectrum of A after immersion in 3 wt.% NaCl ((b)-(d)) Highresolution XPS spectra of Fe 2p3/2, Cr 2p3/2 and O 1s. Figure 5.5(a) presents the normalized XPS spectrum of AB20 high entropy metallic glass alloy before and after immersion in a 3 wt.% NaCl solution. The peaks of Fe 2p, Cr 2p, O 1s and B 1s can be easily seen and the Co 3s and Ni 3p are barely visible. The Fe 2p (Figure 5.5(b)) presents 2 peaks centered at 706.81 and 710.09 eV which correspond to metallic Fe and Fe3+, respectively. In the Cr 2p XPS spectrum (Figure 5.5(c)), the binding energy peaks are centered at 573.53, 575.94, 576.88 eV and can be ascribed to metallic Cr, Cr3+ and Cr6+. 73 Deconvolution of the O 1s peak (Figure 5.5(d)) results again in two component peaks with energies of 529.90 eV that represents a M-O compound characteristic peak corresponding to O-2 species and a second peak (531.87 eV) that represent a M-(O-H)n compound characteristic peak corresponding to O-H. B 1s spectra (Figure 5.5(e)) was also deconvoluted in two peaks which were assigned to a boride (189.82 eV) and B3+ (191.73 eV) bonds respectively. Figure 5.5 (a) Normalised XPS spectrum of AB20 after immersion in 3 wt.% NaCl ((b)-(e)) Highresolution XPS spectra of Fe 2p3/2, Cr 2p3/2, O 1s and B 1s. In Figure 5.6a, the full XPS spectra for sample AB10Si10 clearly shows the coexistence of elemental Fe, Cr, O, and Si, but Co, Ni, and B are indistinguishable from the baseline. The high-resolution XPS spectra of Fe 2p, Cr 2, O 1s, and Si 2p are shown in Figure 5.6b–e. In the 80 81 References 1. Klement, W.; Willens, R.H.; Duwez, P.O.L. Non-Crystalline Structure in Solidified Gold-Silicon Alloys. Nature 1960, 187, 869–870. 2. Miller, M.; Liaw, P. Bulk Metallic Glasses: An Overview; 2008; ISBN 9780387489209. 3. Suryanarayana, C. Rapid Solidification Processing. Encycl. Mater. Sci. Technol. 2002, 1–10, doi:10.1016/B0-08-043152-6/01831-3. 4. Ye, Y.F.; Wang, Q.; Lu, J.; Liu, C.T.; Yang, Y. High-Entropy Alloy: Challenges and Prospects. Mater. Today 2016, 19, 349–362, doi:10.1016/j.mattod.2015.11.026. 5. Inoue, A. Stabilization of Metallic Supercooled Liquid and Bulk Amorphous Alloys. Acta Mater. 2000, 48, 279–306, doi:10.1016/S1359-6454(99)00300-6. 6. Zhang, M.; Wang, A.; Shen, B. Enhancement of Glass-Forming Ability of Fe-Based Bulk Metallic Glasses with High Saturation Magnetic Flux Density. AIP Adv. 2012, 2, doi:10.1063/1.4733340. 7. Akihisa Inoue, Yoshiyuki Shinohara, J.S.G. Thermal and Magnetic Properties of Bulk Fe-Based Glassy Alloys Prepared by Copper Mold Casting. Mater. Trans. JIM 1995, 36, 1427–1433. 8. Louzguine-Luzgin, D. V.; Chen, N.; Churymov, A.Y.; Louzguina-Luzgina, L. V.; Polkin, V.I.; Battezzati, L.; Yavari, A.R. Role of Different Factors in the Glass-Forming Ability of Binary Alloys. J. Mater. Sci. 2015, 50, 1783–1793, doi:10.1007/s10853-0148741-y. 9. Fulchiron, R.; Belyamani, I.; Otaigbe, J.U.; Bounor-Legaré, V. A Simple Method for Tuning the Glass Transition Process in Inorganic Phosphate Glasses. Sci. Rep. 2015, 5, doi:10.1038/srep08369. 10. Turnbull, D. Under What Conditions Can A Glass Be Formed? Contemp. Phys. 1969, 10, 473–488, doi:10.1080/00107516908204405. 11. Shao, G.; Lu, B.; Liu, Y.Q.; Tsakiropoulos, P. Glass Forming Ability of MultiComponent Metallic Systems. Intermetallics 2005, 13, 409–414, doi:10.1016/j.intermet.2004.07.030. 82 12. George, E.P.; Raabe, D.; Ritchie, R.O. High-Entropy Alloys. Nat. Rev. Mater. 2019, 4, 515–534, doi:10.1038/s41578-019-0121-4. 13. Zhang, Y.; Zhou, Y.J. Solid Solution Formation Criteria for High Entropy Alloys. Mater. Sci. Forum 2007, 561–565, 1337–1339, doi:10.4028/www.scientific.net/msf.561565.1337. 14. Guo, S.; Liu, C.T. Phase Stability in High Entropy Alloys: Formation of Solid-Solution Phase or Amorphous Phase. Prog. Nat. Sci. Mater. Int. 2011, 21, 433–446, doi:10.1016/S1002-0071(12)60080-X. 15. Sheikh, S.; Mao, H.; Guo, S. Predicting Solid Solubility in CoCrFeNiMx (M = 4d Transition Metal) High-Entropy Alloys. J. Appl. Phys. 2017, 121, 1–8, doi:10.1063/1.4983762. 16. Liang, X.B.; Wei, M.; Cheng, J.B.; Zhang, W.; Xu, B.S. Reaserch Progress in Advanced Materials of High-Entropy Alloys. Cailiao Gongcheng/Journal Mater. Eng. 2009, 12, 75–79. 17. Sahlberg, M.; Karlsson, D.; Zlotea, C.; Jansson, U. Superior Hydrogen Storage in High Entropy Alloys. Sci. Rep. 2016, 6, 1–6, doi:10.1038/srep36770. 18. Liu, W.H.; Yang, T.; Liu, C.T. Precipitation Hardening in CoCrFeNi-Based High Entropy Alloys. Mater. Chem. Phys. 2018, 210, 2–11, doi:10.1016/j.matchemphys.2017.07.037. 19. Kolano-Burian, A.; Wlodarczyk, P.; Hawelek, L.; Kolano, R.; Polak, M.; Zackiewicz, P.; Temleitner, L. Impact of Cobalt Content on the Crystallization Pattern in the Finemet-Type Ribbons. J. Alloys Compd. 2014, 615, S203–S207, doi:10.1016/j.jallcom.2013.12.066. 20. He, J.Y.; Wang, H.; Wu, Y.; Liu, X.J.; Mao, H.H.; Nieh, T.G.; Lu, Z.P. Precipitation Behavior and Its Effects on Tensile Properties of FeCoNiCr High-Entropy Alloys. Intermetallics 2016, 79, 41–52, doi:10.1016/j.intermet.2016.09.005. 21. Ding, J.; Inoue, A.; Han, Y.; Kong, F.L.; Zhu, S.L.; Wang, Z.; Shalaan, E.; Al-Marzouki, F. High Entropy Effect on Structure and Properties of (Fe,Co,Ni,Cr)-B Amorphous Alloys. J. Alloys Compd. 2017, 696, 345–352, doi:10.1016/j.jallcom.2016.11.223. 22. Qi, T.; Li, Y.; Takeuchi, A.; Xie, G.; Miao, H.; Zhang, W. Soft Magnetic 83 Fe25Co25Ni25(B, Si)25 High Entropy Bulk Metallic Glasses. Intermetallics 2015, 66, 8–12, doi:10.1016/j.intermet.2015.06.015. 23. B. Cullity Elements of X-Ray Diffraction; 3rd ed.; 1967; 24. Nagaraj, S.K.; Shivanna, S.; Subramani, N.K. Revisiting Powder X-Ray Diffraction Technique : A Powerful Tool to Characterize Polymers and Their Composite Films. 2016, 4, 1–5, doi:10.4172/2321-6212.1000158. 25. Ho hne G W Hemminger and H.-J Flammersheim Differential Scanning Calorimetry : An Introduction for Practitioners; Berlin: Springer-Verlag, 1996; 26. W. ZhouZ.L. Wang Scanning Microscopy for Nanotechnology. Techniques and Applications; 2007; ISBN 978-1-4419-2209-0. 27. Gütlich, P.; Link, R.; Trautwein, A. Mössbauer Spectroscopy and Transition Metal Chemistry: Fundamentals and Applications; 1978; Vol. 3; ISBN 9783540884279. 28. Brand, R. A., Lauer, J. and Herlach, D.M. No Title. J. Phys. F Met. Phys 1983, 13, 875. 29. Oliver, W.C.; Pharr, G.M. An Improved Technique for Determining Hardness and Elastic Modulus Using Load and Displacement Sensing Indentation Experiments. J. Mater. Res. 1992, 7, 1564–1583, doi:10.1557/jmr.1992.1564. 30. MariAnne Sullivan Measuring, Evaluating, and Describing Pile-Up and Sink-In During Nanoindentation of Thin Films on Substrates. J. Chem. Inf. Model. 2013, 53, 1689–1699. 31. Sullivan, M.A.; Prorok, B.C. Newly Discovered Pile up Effects during Nanoindentation. Conf. Proc. Soc. Exp. Mech. Ser. 2015, 8, 1–5, doi:10.1007/978-3-319-07004-9_1. 32. Zhang, L.; Name, P.; Hora, P.; Name, P.; Jang, J.; Name, P.; Kirkes, L.; Name, P.; Miller, C.; Name, P.; et al. Electrochemical Tests Under TP 06-02 Effective. 2021, 1–10. 33. Poursaee, A. Corrosion Sensing for Assessing and Monitoring Civil Infrastructures; Woodhead Publishing Limited, 2014; Vol. 1; ISBN 9780857094322. 34. Kruger, J. The Oxide Films Formed on Copper Single Crystal Surfaces in Water. J. Electrochem. Soc. 1959, 108, 503, doi:10.1149/1.2428124. 35. Kong, K.H.; Kim, K.C.; Kim, W.T.; Kim, D.H. Microstructural Features of Multicomponent FeCoCrNiSi x Alloys . Appl. Microsc. 2015, 45, 32–36, doi:10.9729/am.2015.45.1.32. 84 36. Xing, Q.W.; Zhang, Y. Amorphous Phase Formation Rules in High-Entropy Alloys. Chinese Phys. B 2017, 26, doi:10.1088/1674-1056/26/1/018104. 37. Zhang, Y.; Zhou, Y.J.; Lin, J.P.; Chen, G.L.; Liaw, P.K. Solid-Solution Phase Formation Rules for Multi-Component Alloys. Adv. Eng. Mater. 2008, 10, 534–538, doi:10.1002/adem.200700240. 38. Załuska, A.; Matyja, H. Crystallization Characteristics of Amorphous Fe-Si-B Alloys. J. Mater. Sci. 1983, 18, 2163–2172, doi:10.1007/BF00555011. 39. Shpak, A.P.; Il’Inskii, A.G.; Marunyak, A. V.; Slukhovskyy, O.I.; Lepeeva, Y. V.; Dekhtyar, A.; Kaban, I.; Mattern, N.; Eckert, J. Crystallization of Fe82Si2B16 and Fe82Si4B14 Metallic Glasses upon Isothermal and Non-Isothermal Annealing. EPJ Web Conf. 2011, 15, 14–17, doi:10.1051/epjconf/20111501008. 40. Cheng, Y.T.; Cheng, C.M. Relationships between Hardness, Elastic Modulus, and the Work of Indentation. Appl. Phys. Lett. 1998, 73, 614–616, doi:10.1063/1.121873. 41. Pellicer, E.; Pané, S.; Sivaraman, K.M.; Ergeneman, O.; Suriñach, S.; Baró, M.D.; Nelson, B.J.; Sort, J. Effects of the Anion in Glycine-Containing Electrolytes on the Mechanical Properties of Electrodeposited Co-Ni Films. Mater. Chem. Phys. 2011, 130, 1380–1386, doi:10.1016/j.matchemphys.2011.09.032. 42. Wang, X.F.; Yang, X.P.; Guo, Z.D.; Zhou, Y.C.; Song, H.W. Nanoindentation Characterization of Mechanical Properties of Ferrite and Austenite in Duplex Stainless Steel. Adv. Mater. Res. 2007, 26–28, 1165–1170, doi:10.4028/www.scientific.net/amr.26-28.1165. 43. Guo, W.; Choi, P.P.; Seol, J.B. Amorphous Phase Separation in an Fe-Based Bulk Metallic Glass. Mater. Lett. 2017, 190, 161–164, doi:10.1016/j.matlet.2017.01.012. 44. Szlufarska, I.; Kalia, R.K.; Nakano, A.; Vashishta, P. Atomistic Mechanisms of Amorphization during Nanoindentation of SiC: A Molecular Dynamics Study. Phys. Rev. B - Condens. Matter Mater. Phys. 2005, 71, 1–11, doi:10.1103/PhysRevB.71.174113. 45. Sinha, S.; Mirshams, R.A.; Wang, T.; Nene, S.S.; Frank, M.; Liu, K.; Mishra, R.S. Nanoindentation Behavior of High Entropy Alloys with Transformation-Induced Plasticity. Sci. Rep. 2019, 9, 1–11, doi:10.1038/s41598-019-43174-x. 85 46. Sun, Y.; Chen, P.; Liu, L.; Yan, M.; Wu, X.; Yu, C.; Liu, Z. Local Mechanical Properties of AlxCoCrCuFeNi High Entropy Alloy Characterized Using Nanoindentation. Intermetallics 2018, 93, 85–88, doi:10.1016/j.intermet.2017.11.010. 47. Duan, F.H.; Pan, J.; Lin, Y.; Li, Y. Significant Structural Relaxation in a Mo[Sbnd]O Binary Amorphous Alloy. J. Non. Cryst. Solids 2019, 514, 10–14, doi:10.1016/j.jnoncrysol.2019.03.038. 48. Gludovatz, B.; George, E.P.; Ritchie, R.O. Processing, Microstructure and Mechanical Properties of the CrMnFeCoNi High-Entropy Alloy. Jom 2015, 67, 2262–2270, doi:10.1007/s11837-015-1589-z. 49. Leitner, A.; Maier-Kiener, V.; Kiener, D. Extraction of Flow Behavior and Hall–Petch Parameters Using a Nanoindentation Multiple Sharp Tip Approach. Adv. Eng. Mater. 2017, 19, 1–9, doi:10.1002/adem.201600669. 50. Attaf, M.T. Connection between the Loading Curve Models in Elastoplastic Indentation. Mater. Lett. 2004, 58, 3491–3498, doi:10.1016/j.matlet.2004.06.049. 51. Duan, F.H.; Pan, J.; Lin, Y.; Li, Y. Significant Structural Relaxation in a Mo[Sbnd]O Binary Amorphous Alloy. J. Non. Cryst. Solids 2019, 514, 10–14, doi:10.1016/j.jnoncrysol.2019.03.038. 52. Aliaga, L.C.R.; Beringues, J.F.; Suriñach, S.; Baró, M.D.; Kiminami, C.S.; Bolfarini, C.; Botta, W.J.; Viñas, J.S. Comparative Study of Nanoindentation on Melt-Spun Ribbon and Bulk Metallic Glass with Ni60Nb37B3 Composition. J. Mater. Res. 2013, 28, 2740–2746, doi:10.1557/jmr.2013.260. 53. Zhang, L.S. Preparation of CoCrFeNiCuMnSix High Entropy Alloys and Their Microstructure and Properties. Adv. Mater. Res. 2013, 750–752, 615–618, doi:10.4028/www.scientific.net/AMR.750-752.615. 54. Zhang, Z. Elastic Properties of Bulk-Metallic Glasses Studied by Resonant Ultrasound Spectroscopy. 2008. 55. Fauth, F.; Peral, I.; Popescu, C.; Knapp, M. The New Material Science Powder Diffraction Beamline at ALBA Synchrotron. Powder Diffr. 2013, 28, 360–370. 56. Schneider, C.A.; Rasband, W.S.; Eliceiri, K.W. NIH Image to ImageJ: 25 Years of Image Analysis. Nat. Methods 2012, 9, 671–675, doi:10.1038/nmeth.2089. 86 57. Zhang, H.; Zhong, X.C.; He, Y.Z.; Li, W.H.; Wu, W.F.; Chen, G.; Guo, S. Effect of High Configuration Entropy and Rare Earth Addition on Boride Precipitation and Mechanical Properties of Multi-Principal-Element Alloys. J. Mater. Eng. Perform. 2017, 26, 3750–3755, doi:10.1007/s11665-017-2831-3. 58. Leyland, A.; Matthews, A. On the Significance of the H/E Ratio in Wear Control: A Nanocomposite Coating Approach to Optimised Tribological Behaviour. Wear 2000, 246, 1–11, doi:10.1016/S0043-1648(00)00488-9. 59. Fallis, A.. Measuring, Evaluating, and Describing Pile-Up and Sink-In During Nanoindentation of Thin Films on Substrates. J. Chem. Inf. Model. 2013, 53, 1689–1699. 60. Mirshams, R.A.; Srivastava, A.K. Effect of Pile-up on Nanoindentation Measurements of Polycrystalline Bulk Metals. Adv. Mater. Res. 2014, 853, 143–150, doi:10.4028/www.scientific.net/AMR.853.143. 61. Beegan, D.; Chowdhury, S.; Laugier, M.T. The Nanoindentation Behaviour of Hard and Soft Films on Silicon Substrates. Thin Solid Films 2004, 466, 167–174, doi:10.1016/j.tsf.2004.03.006. 62. Burik, P.; Pešek, L.; Voleský, L. Effect of Pile-up on the Mechanical Characteristics of Steel with Different Strain History by Depth Sensing Indentation. Met. 2014 - 23rd Int. Conf. Metall. Mater. Conf. Proc. 2014, 2, 629–633. 63. Oliver, W.C.; Pharr, G.M. Measurement of Hardness and Elastic Modulus by Instrumented Indentation: Advances in Understanding and Refinements to Methodology. J. Mater. Res. 2004, 19, 3–20, doi:10.1557/jmr.2004.19.1.3. 64. Koch, G.H.; Brongers, M.P.H.; Thompson, N.G.; Virmani, Y.P.; Payer, J.. Corrosion Cost and Preventive Strategies in the United State. Natl. Tech. Inf. Serv. Rep. No. FHWA-RD-01-156 2001. 65. H.H.Liebermann(ed.) Rapidly Solidified Alloys Processes-Structures-PropertiesApplications; Marcel Dekker, New York, NY, 1993; ISBN 9780824789510. 66. Wang, S. Corrosion Resistance and Electrocatalytic Properties of Metallic Glasses. Met. Glas. - Form. Prop. 2016, doi:10.5772/63677. 67. Brien, V.; Khare, V.; Herbst, F.; Weisbecker, P.; Ledeuil, J.B.; de Weerd, M.C.; Machizaud, F.; Dubois, J.M. Influence of Boron Content on the Microstructure of 87 Sintered Al62.5-XCu25.3Fe12.2Bx Alloys (x = 0, 3, 5). J. Mater. Res. 2004, 19, 2974– 2980. 68. Zhang, J.S.; Xue, Y.J.; Guo, Y.J.; Xu, C.X.; Liang, W. Effect of Si on As-Cast Microstructure in Quasicrystalline Al-Cu -Fe Alloy. Mater. Sci. Forum 2007, 546–549, 619–622, doi:10.4028/www.scientific.net/msf.546-549.619. 69. Dutta, R.S.; Dey, G.K. Effects of Partial Crystallinity and Quenched-in Defects on Corrosion of a Rapidly Solidified Ti-Cu Alloy. Bull. Mater. Sci. 2003, 26, 477–482, doi:10.1007/BF02707344. 70. Jayaraj, J.; Kim, Y.C.; Kim, K.B.; Seok, H.K.; Fleury, E. Corrosion Studies on Fe-Based Amorphous Alloys in Simulated PEM Fuel Cell Environment. Sci. Technol. Adv. Mater. 2005, 6, 282–289, doi:10.1016/j.stam.2005.02.019. 71. Zander, D.; Heisterkamp, B.; Gallino, I. Corrosion Resistance of Cu-Zr-Al-Y and ZrCu-Ni-Al-Nb Bulk Metallic Glasses. J. Alloys Compd. 2007, 434–435, 234–236, doi:10.1016/j.jallcom.2006.08.112. 72. Raicheff, R.; Zaprianova, V. EFFECT OF COBALT AND NICKEL ALLOYING ON CORROSION BEHAVIOUR OF AMORPHOUS Fe-B-Si ALLOYS. J. Univ. Chem. Technol. Metall. 2009, 44, 61–65. 73. Qin, C.; Zhang, W.; Asami, K.; Ohtsu, N.; Inoue, A. Glass Formation, Corrosion Behavior and Mechanical Properties of Bulk Glassy Cu–Hf–Ti–Nb Alloys. Acta Mater. 2005, 53, 3903–3911, doi:10.1016/j.actamat.2005.04.037. 74. Li, M.; Chen, Q.; Cui, X.; Peng, X.; Huang, G. Evaluation of Corrosion Resistance of the Single-Phase Light Refractory High Entropy Alloy TiCrVNb0.5Al0.5 in Chloride Environment. J. Alloys Compd. 2021, 857, doi:10.1016/j.jallcom.2020.158278. 75. Pang, S.J.; Zhang, T.; Asami, K.; Inoue, A. Synthesis of Fe-Cr-Mo-C-B-P Bulk Metallic Glasses with High Corrosion Resistance. Acta Mater. 2002, 50, doi:10.1016/S13596454(01)00366-4. 76. Shang, X.L.; Wang, Z.J.; Wu, Q.F.; Wang, J.C.; Li, J.J.; Yu, J.K. Effect of Mo Addition on Corrosion Behavior of High-Entropy Alloys CoCrFeNiMo x in Aqueous Environments. Acta Metall. Sin. (English Lett. 2019, 32, 41–51, doi:10.1007/s40195018-0812-7. 88 77. Bijalwan, P.; Kumar, A.; Nayak, S.K.; Banerjee, A.; Dutta, M.; Laha, T. Microstructure and Corrosion Behavior of Fe-Based Amorphous Composite Coatings Developed by Atmospheric Plasma Spraying. J. Alloys Compd. 2019, 796, doi:10.1016/j.jallcom.2019.05.046. 78. Guo, W.; Liu, B. Microstructure and Corrosion Behavior of Laser Cladding FeCoNiCrBSi Based High-Entropy Alloy Coatings. Coatings 2022, 12, 1–18. 79. Huang, F.; Kang, J. jie; Yue, W.; Fu, Z. qiang; Zhu, L. na; She, D. shun; Liang, J.; Wang, C. biao Corrosion Behavior of FeCrMoCBY Amorphous Coating Fabricated by HighVelocity Air Fuel Spraying. J. Therm. Spray Technol. 2019, 28, 842–850, doi:10.1007/s11666-019-00843-7. 80. Cui, P.; Bao, Z.; Liu, Y.; Zhou, F.; Lai, Z.; Zhou, Y.; Zhu, J. Corrosion Behavior and Mechanism of Dual Phase Fe1.125Ni1.06CrAl High Entropy Alloy. Corros. Sci. 2022, 201, 110276, doi:10.1016/j.corsci.2022.110276. 81. Li, Y.; Wang, S.; Wang, X.; Yin, M.; Zhang, W. New FeNiCrMo(P, C, B) High-Entropy Bulk Metallic Glasses with Unusual Thermal Stability and Corrosion Resistance. J. Mater. Sci. Technol. 2020, 43, 32–39, doi:10.1016/j.jmst.2020.01.020. 82. Gong, P.; Wang, D.; Zhang, C.; Wang, Y.; Jamili-Shirvan, Z.; Yao, K.; Wang, X. Corrosion Behavior of TiZrHfBeCu(Ni) High-Entropy Bulk Metallic Glasses in 3.5 Wt. % NaCl. npj Mater. Degrad. 2022, 6, 1–14, doi:10.1038/s41529-022-00287-5. 83. Takeuchi, A.; Amiya, K.; Wada, T.; Yubuta, K.; Zhang, W.; Makino, A. Entropies in Alloy Design for High-Entropy and Bulk Glassy Alloys. Entropy 2013, 15, doi:10.3390/e15093810. 84. Panahi, S.L.; Garcia-Ramón, M.; Pineda, E.; Bruna, P. New (FeCoCrNi)-(B,Si) HighEntropy Metallic Glasses, Study of the Crystallization Processes by X-Ray Diffraction and Mössbauer Spectroscopy. J. Non. Cryst. Solids 2020, 547, 120301, doi:10.1016/J.JNONCRYSOL.2020.120301. 85. Kumar, N.; Fusco, M.; Komarasamy, M.; Mishra, R.S.; Bourham, M.; Murty, K.L. Understanding Effect of 3.5 Wt.% NaCl on the Corrosion of Al0.1CoCrFeNi HighEntropy Alloy. J. Nucl. Mater. 2017, 495, 154–163, doi:10.1016/j.jnucmat.2017.08.015. 86. Córdoba-Torres, P. Relationship between Constant-Phase Element (CPE) Parameters 89 and Physical Properties of Films with a Distributed Resistivity. Electrochim. Acta 2017, 225, doi:10.1016/j.electacta.2016.12.087. 87. Ulum, M.F.; Caesarendra, W.; Alavi, R.; Hermawan, H. In-Vivo Corrosion Characterization and Assessment of Absorbable Metal Implants. Coatings 2019, 9. 88. Sun, Y.P.; Wang, Z.; Yang, H.J.; Lan, A.D.; Qiao, J.W. Effects of the Element La on the Corrosion Properties of CrMnFeNi High Entropy Alloys. J. Alloys Compd. 2020, 842, doi:10.1016/j.jallcom.2020.155825. 89. Bar-Cohen, Y. High Temperature Materials and Mechanisms; CRC Press, Boca Raton, 2017; ISBN 9781466566460. 90. Beke, D.L.; Erdélyi, G. On the Diffusion in High-Entropy Alloys. Mater. Lett. 2016, 164, doi:10.1016/j.matlet.2015.09.028. 91. Bredar, A.R.C.; Chown, A.L.; Burton, A.R.; Farnum, B.H. Electrochemical Impedance Spectroscopy of Metal Oxide Electrodes for Energy Applications. ACS Appl. Energy Mater. 2020, 3. 92. Qiu, Y.; Thomas, S.; Gibson, M.A.; Fraser, H.L.; Pohl, K.; Birbilis, N. Microstructure and Corrosion Properties of the Low-Density Single-Phase Compositionally Complex Alloy AlTiVCr. Corros. Sci. 2018, 133, doi:10.1016/j.corsci.2018.01.035. 93. Qiu, Y.; Thomas, S.; Fabijanic, D.; Barlow, A.J.; Fraser, H.L.; Birbilis, N. Microstructural Evolution, Electrochemical and Corrosion Properties of Al x CoCrFeNiTi y High Entropy Alloys. Mater. Des. 2019, 170, doi:10.1016/j.matdes.2019.107698. 94. Koga, G.Y.; Otani, L.B.; Silva, A.M.B.; Roche, V.; Nogueira, R.P.; Jorge, A.M.; Bolfarini, C.; Kiminami, C.S.; Botta, W.J. Materials Characterization and Corrosion Resistance of Boron-Containing-Austenitic Stainless Steels Produced by Rapid Solidification Techniques., doi:10.3390/ma11112189. 95. Masumoto, T.; Hashimoto, K.; Masumoto, T.; Corrosion, K.H.; Of, P.; Metals, A. CORROSION PROPERTIES OF AMORPHOUS METALS To Cite This Version : HAL Id : Jpa-00220328. 1980. 96. Muangtong, P.; Rodchanarowan, A.; Chaysuwan, D.; Chanlek, N.; Goodall, R. The Corrosion Behaviour of CoCrFeNi-x (X = Cu, Al, Sn) High Entropy Alloy Systems in