Understanding chemical ordering in near-equiatomic bulk in partial fu - PDF document

Understanding chemical ordering in near-
Understanding chemical ordering in near-equiatomic bulk in partial fulfillment of the requirements ACKNOWLEDGMENTS There are so many people that I would like to thank for helping me in one way or another to get to this point. First, I would like to thank my advisor Dr. Laura Lewis for giving me the opportunity to work with her in this amazing project and for her constant support throughout these years. Her mentorship has been extremely valuable to me, and has allowed me to grow both at a personal and professional level. I am so grateful for her guidance, for being always available to share her knowledge and experience, for her Thank you so much Dr. Ando for all your guidance and help throughout these years. I literally wouldn’t have come to Northeastern if it weren’t for you. I have learned so much from you and your amazing classes, and I really appreciate your contributions as part of my dissertation committee. take a part in my dissertation committee. Your time, interest and insiI would like to thank collaborators from other institutions with whom I have had the pleasure of working with; Dr. Barmak at Columbia University, for always being aring her wealth of knowledge in metallurgy and magnetism. Late Distinguished Professor Emeritus Dr. Joseph I. Goldstein for his invaluable guidance and contributions to the FeNi project, particularly regarding the understanding of meteorites. Dr. Stephens at the NSLS (B

NL) for all his help with synchrotron X-
NL) for all his help with synchrotron X-ray experiments. Dr. Knight, Dr. Langridge and Dr. Fortes at Rutherford Appleton Lab for their help with neutron diffraction experiments and for helpful discussions. Dr. Kramer, Matthew Besser, Kevin Dennis, Larry Jones and Brandt Jensen at Ames Lab for all their help with I would like to express my gratitude to all the people I have worked with at the Nanomagnetism Lab; I have learned something from each and every one of you. Thank you Professor Heiman, Felix, Barbara, Luke, h, Nina, Ian, Brian, Mahboobeh, Michelle, Brad, Nicholas, Piers, Jake, Drew, and Juan Sebastián. Special thanks to: Luke, for all his help in the FeNi project and his insightful advice in scientific, career and personal aspects; Nina, for teaching me how to operate every single piece of equipment in the lab, for so many discussions about L1 systems and life, and for being an amazing friend; Ian, for making the grad-student experience much more enjoyable. I wish to thank Pat, Jessica and Rob at the Chemical Engineering Department, as well as Noah and Joyce at the Mechanical and Industrial Engineering Department, for always being willing to help me in every possible way. I would like to acknowledge that the research presented in this dissertation was ce Foundation (CMMI Division, Grant No. 1129433), by Rogers Corporation, and by Northeastern University. Also, I am grateful for the generous Dissertati

on Completion Fellowship from the Office
on Completion Fellowship from the Office of the Provost that supported me through my last semester. Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-98CH10886. Experiments at the ISIS Pulsed Neutron and Muon Source were supported by a beamtime allocation from the U.K. Science and Technology Facilities Council. Northeastern University is gratefully acknowledged for providing the isotope used in this Last but not least, to my family, for their unconditional love and constant encouragement throughout my life. To my wonderful parents, for going above and beyond to give me the best opportunities in life, for setting an example of discipline and dedication to accomplish whatever goals I set for myself, for encouraging me to pursue my dreams. All I am and all I have become I owe it to you both. To my siblings, for always being there for me. You are such an important part of my life, and I am extremely grateful for having you as my family. I really admire and love you all so much. To my life has brought so much joy and happiness to mine. To my husband, Germán, for believing in me more than anyone else. You have given up so much to make my career a priority, and I am so grateful for that. It was your endless and selfless support that got me thThe development of new

materials to facilitate technological ad
materials to facilitate technological advancements relies on the understanding of structure-processing-property correlations. This is particularly true in the field of permanent magnetic materials, in which reconciliation between the electronic/crystal/micro-structure of matter and the resultant magnetic properties has allowed enormous progress in the last 100 years. Permanent magnets, used for energy conversion and generation, functionalize much of modern technology. The growing demand for advanced permanent magnets, coupled with recent limited availability and price volatility of the rare-earth elements that constitute them, has resulted in an increased interest from the scientific and tecrealize novel magnetic materials made of earth-abundant elements. This dissertation examines fundamental aspects of ferromagnetic materials with the tetragonal, chemically ordered L1 crystal structure, considered as candidate materials for next-generation permanent magnets. Understanding of L1-structured compounds and the chemical order-disorder phase transformation by which they form is of fundamental importance for the successful integration of these materials into technological applications. This dissertation examines and quantifies the effects of intrinsic and extrinsic parameter variation on thermodynamic and kinetic aspects of the chemical order-disorder phase transformation in near-equiatomic FePd and FeNi, as well as

the structure-processing-property corre
the structure-processing-property correlations associated with these systems. FePd was used as a model system to evaluate the effects of ternary alloying additions and of severe plastic deformation on the chemical order-disorder transformation using specific structural, magnetic and calorimetric probes. Results obtained from this model system verify the potential of these processing protocols as metthe barriers to the formation of the L1 phase. These protocols were then applied toward -type FeNi phase. L1 FeNi is only found naturally in meteorites that form over periods of billions of years, and bulk laboratory synthesis has not been accomplished to date. In this dissertation processed FeNi-based alloys characterized by high-resolution neutron diffraction were confirmed to exhibit an unreported tetragonal chemically disordered phase, hypothesized to be a precursor to the L1 FeNi phase. These results furnish new knowledge regarding the Fe-Ni system and the mechanism of the chemical order-disorder transformation in ferrous compounds, with relevant implications TABLE OF CONTENTS LIST OF FIGURES ............................................................................................................ xLIST OF TABLES ............................................................................................................ xxINTRODUCTION ..........................................................................

........................ 1Motivation ...
........................ 1Motivation ....................................................................................................... 1Overview of research ....................................................................................... 4Organization of the dissertation ...................................................................... 6FUNDAMENTALS ................................................................................................ 8Literature review ............................................................................................. 8Chemically order in binary alloys and the L1 crystal structure ......... 9Overview of the Iron-Palladium and Iron-Nickel material systems .. 13-type chemical order .................................................................... 23Factors that influence the A1 phase transformation in ferrous alloys ................................................................................................. 37Formalism ...................................................................................................... 43Thermodynamics of Phase Transformations ..................................... 44Kinetics of Phase Transformations .................................................... 49Key Concepts in Magnetism ............................................................. 55EXPERIMENTAL TECHNIQUES ....................................................

.................. 63Sample synthesis ..
.................. 63Sample synthesis ........................................................................................... 66Synthesis of bulk FePd and FeNi-based alloys via arc-melting ........ 67Synthesis of bulk FePd and FeNi-based alloys via drop-casting ...... 68Sample processing ......................................................................................... 68 facilitate further processing ..... 68Processing of FeNi-based alloys through cryomilling ...................... 69 alloys through cold-rolling ...... 70Annealing of FePd and FeNi-based alloys to promote phase transformations .................................................................................. 71Characterization methods .............................................................................. 72Compositional analysis of bulk alloys ............................................... 72-ray and neutron diffraction ............... 73Magnetic characterization: magnetometry ........................................ 81Differential scanning calorimetry: analysis of phase transformation RESULTS AND DISCUSSION ........................................................................... 88Aim 1: Understanding the effects of intrinsic modification, ternary alloying additions, on the phase transformation character, structure and magnetic properties of L1 FePd ................................................................... 88I

ntroduction and justification ..........
ntroduction and justification ............................................................ 89FePdM sample synthesis and characterization .................................. 90FePd and FePdM (M = Ni or Cu) Results ......................................... 92Discussion of results obtained for FePd and FePdM (M = Ni or Cu) 111Significance of work on composition-in FePd ............................................................................................. 124Aim 2. Understanding the effects of extrinsic modification, plastic deformation, on the order-disorder A1 phase transformation in FePd 126Introduction and justification .......................................................... 126FePd sample synthesis, processing and characterization ................ 127Results from cold-rolled FePd ......................................................... 128Discussion from results obtained for cold-rolled FePd ................... 142ructure correlations in FePd 149Aim 3. Evaluate the effects of ternary alloying additions and plastic deformation on the A1 transformation of FeNi. ................................ 152Introduction and justification .......................................................... 152FeNi-based sample synthesis, processing and characterization ...... 155FeNi and FeNiM (M = Al, Ti, V) Results ....................................... 158Discussion of results obtained for FeNi and FeNiM (M = Al, Ti, V)

171Significance of work on compositicor
171Significance of work on compositicorrelations in FeNi ......................................................................... 180SUMMARY AND CONCLUSIONS ................................................................. 182RECOMMENDATIONS FOR FUTURE WORK ............................................. 188NOMENCLATURE ........................................................................................... 195REFERENCES ................................................................................................... 199APPENDICES .................................................................................................... 215Appendix 1 .......................................................................................................... 217Appendix 2 .......................................................................................................... 220Appendix 3 .......................................................................................................... 226Appendix 4 .......................................................................................................... 229Figure 1. Schematic representation of a) fcc (A1), and b) L1 crystal structures in binary AB alloys, where A and B are different elements. Half-shaded spheres represent an equal probability of lattice site occupancy by A or B atoms. For structure, planes along the -axis have been shaded to emphasize t

he chemical order. .....................
he chemical order. .................................................................................................. 3Figure 2. Common chemically ordered lattices: a) B2, b) L1, c) L1, d) D0, e) C0Image adapted from [27]. ................................................................................ 10Figure 3. Schematic representation of a) fcc (A1), and b) L1 crystal structures, where A and B are constituent elements. Half-shaded spheres represent an equal probability of lattice site occupancy by A or B atoms. The tetragonal structure is shown with the same orientation of thcell (fct representation), and dashed lines are used to delineate the primitive standard bct unit cell. ...................................................................................... 11Figure 4. Crystallographic domains found upon formation of an L1 phase. a) Translational domains, separated by an anti-phase boundary. b) Orientational domains, for which the tetragonal axis is directed along any of the three different basis vector of the cubic cell. These domains are usually called variants. ........................................................................................................... 12Figure 5. Binary Iron-Palladium phase diagram. Image adapted from [32]. .................... 14 curves for near-equiatomic FePd. L1 volume fractions of 5, 20, 50 and 90% are presented as a function of temperature (Celsius) and time (h)

of isothermal annealing. Adapted from [
of isothermal annealing. Adapted from [39]. ...................................................... 15Figure 7. Schematic representation of the A1 transformation in FePd through the intermediate A6 phase suggested by Vlasova et al. [46]. ............................... 16Figure 8. Fe-Ni low temperature phase diagram with a bcc phase (), fcc phase (), L1ordered FeNi) phase. Image adapted from [52]. ................................................................................................................. 19Figure 9. Magnetic contribution to the free energy of a) vacancy formation, and b) vacancy migration, in fcc FeNi alloys as a function of temperature and composition. The Curie temperature () is the critical temperature at which there is a ferromagnetic to paramagnetic transition. Image adapted from [69].Figure 10. Magnetic contribution to the free energy for diffusion in fcc FeCurie temperature () is the critical temperature at which there is a ferromagnetic to paramagnetic transition. Image adapted from [69]. ............ 21Figure 11. Variation of parameter with temperature for two different chemically ordered alloys. Image adapted from [80]. ....................................................... 25Figure 12. Calculated X-ray diffraction patterns using the FullProf Suite [82] for: a) A1 FePd, and b) L1 FePd. Asterisks designate superlattice peaks, while arrows show peak-splitting. ...................

........................................
..................................................................... 26Figure 13. Typical microstructures resultant from the A1 transformation. a) Tweed contrast, b) Micro-twins and anti-phase boundaries, c) Poly-twinned microstructure. Images adapted from [41]. ..................................................... 27Figure 14. Schematic representation that shows mutually orthogonal nuclei of L1) bands. Image adapted from [41]. ................................... 28Figure 15. Schematic representation of the magnetic domain configuration in L1 poly-twinned alloys. Thin lines represent micro-domain boundaries, thick lines are macro-domain boundaries and serrated lines are mobile domain-walls cutting across the array of micro-twins. Arrows represent magnetization vectors. Image adapted from [42] ................................................................................. 29Figure 16. Simulated a) X-ray and b) neutron diffraction patterns for L1 FeNi considering lattice parameters reporteters reportepeaks (marked by asterisks) have intensities near the signal-to-noise limit, and there is no evident peak-splitting, except at very high angles in X-ray ) peak). The X-ray diffraction pattern was calculated using the FullProf Suite [82], while the neutron diffraction pattern was simulated by collaborators at Rutherford Appleton Laboratories. .......... 31 FeNi, simulated by collaborators at Rutherford Appleton Lab

oratory, considering lattice parameters
oratory, considering lattice parameters reported by Albertsen [62]. In contrast to the diffraction pattern for L1 FeNi made from FeNi is dominated by superlattice peaks, and the intensity of the fundamental Bragg reflections is greatly decreased due to the near null-scattering combination of Fe and neutron scattering lengths. .............................................................................. 33Figure 18. Isothermal sections of Fe-Ni-Pd phase diagram at a) 650 °C b) 590 °C and c) 550 °C, as proposed by Horiuchi et al. [110] from phenomenological calculations. ...................................................................... 39Figure 19. a) Microstructure of fully L1 ordered FePd alloy, obtained by annealing at 500 °C for 100 h. b) Microstructure of fully L1 ordered FePd alloy, obtained by a cold-work reduction of 97% followed by annealing at 500 °C for 6 h. Image adapted from [35]. ................................................................................ 40Figure 20. Electrical resistance measured on isothermal annealing of FePd at 843 K with or without a compressive stress of 40 MPa. A change in the electrical resistance represents the chemical disorder-order transformation. Image adapted from [115]. ......................................................................................... 41Figure 21. Magnetization curves for and directions of a single crystal of Fe-55 at% Pd after a)

ordering at 673 K for 1 h, and b) succe
ordering at 673 K for 1 h, and b) successive ordering at 773 K for 24 h. Image adapted from [117]. ..................................................................... 42Figure 22. Magnetization curves for x, y and z directions of a single crystal of Fe-55 at% Pd after a) ordering at 673 K for 1h under a magnetic field of 10 T, and b) succesive ordering at 773 K for 24 h. Image adapted from [117]. ................. 43Figure 23. Variation of Gibbs free energy with configuration, showing equilibrium states = 0. The stable state is A, while B is a metastable state. Image adapted from [27]. ........................................................................................................ 46Figure 24. Thermodynamic characteristics of a) first-order phase transformations and b) second-order phase transformations. Image adapted from [27]. ..................... 48Figure 25. Variation of Gibbs free energy with configurati for the transformation BA, and the activation energy that should be overcome for the transformation to occur. Image adapted from [27]. ............ 49Figure 26. Characteristic sigmoidal curve associated with isothermal nucleation and growth processes. Image adapted from [66]. .................................................. 54) curve of a ferromagnet, showing saturation magnetization remanence , and coercivity . Magnetic domains at different stages of the [132]. b) Hysteresis loops for soft

and hard ferromagnets. Image adapted fro
and hard ferromagnets. Image adapted from [134] ................................................................................................................ 58ation results of the variation in the magnetocrystalline anisotropy energy of L1 type FePt as a function of a) axial ratio and long-range order parameter [141], and b) long-range order parameter and Fe content [142]. .................................................................................................. 60Figure 29. Permanent magnet performance, as quantified by the energy density in MGOe, versus average selling price per kg. Sintered rare earth based magnets MGOe. Image adapted from [146]. .. 62Figure 30. Diffraction of X-rays by a crystal. For X-rays to interfdistance ABC needs to be an integer number of the wavelength of the X-rays. From geometric construction, the distance ABC is equal to 2sin. .............. 75Figure 31. Comparison of X-ray diffraction under ideal conditions and with microstrain. Image adapted from [160] ............................................................................... 76Figure 32. Schematic representation of a Vibrating Sample Magnetometer (VSM). The oscillating motion is achieved by using a linear motor. Image adapted from [133]. ............................................................................................................... 82Figure 33. Schematic representation of Simultaneous Thermal

Analysis (STA). Adapted from [171]. ....
Analysis (STA). Adapted from [171]. ...................................................................................................... 85Figure 34. X-ray diffractograms obtained for binary FePd in the a) as-arc-melted state, and b) annealed state. Reference patterns of fcc FePd (JCPDS 04-003-5130 c FePd (JCPDS 04-003-5130 )0 FePd (JCPDS 03-065-9971 [48]) are included for comparison.Figure 35. Modified Williamson-Hall plot for FePd in its as-arc-melted (blue, open circles) and annealed (red, closed squares) state. The broken lines represent the linear fit to the data, from which the intercept and slope are extracted to estimate the crystallite size and microstrain. .................................................. 95Figure 36. Magnetization as a function of applied field measured at room temperature for FePd in its as-arc-melted (blue) and annealed (red) states. ............................ 95Figure 37. Differential scanning calorimetry results at 20 K/min on FePd in the as-arc-melted (blue, top) and annealed (red, bottom) states. For the annealed sample, consecutive heating scans are labeled “1 scan”. ................... 96Figure 38. X-ray diffractograms obtained for ternary FePdNi samples in their a) as-arc-melted state, and b) annealed state. Samples are identified by their determined Ni content. Diffractograms obtained for binary as-arc-melted and annealed FePd ( = 0) are included for comparison, as well

as reference patterns for fcc arison, a
as reference patterns for fcc arison, as well as reference patterns for fcc )0 (JCPDS 03-065-9971 [48]). ....... 99Figure 39. Multi-peak structures in the vicinity of the () diffraction peak for a) = 5.2 sample and b) annealed = 7.5 sample. Peaks labeled with fcc phase(s). ................................................................... 99Figure 40. Variation with Ni content of unit cell volume, ratio, parameter, crystallite size () and microstrain () of the fcc (as-arc-melted) and L1(annealed) phases. Results for binary FePd ( = 0) have been included for comparison. Lines are drawn to guide the eye. ............................................. 101Figure 41. Magnetization as a function of applied field measured at room temperature for ternary FePdNi samples in their a) as-arc-melted, and b) annealed states. Samples are identified by their determined Ni content. for binary as-arc-melted and annealed FePd ( = 0) are included for comparison. ................................................................................................... 102Figure 42. Evolution of with Ni content for as-arc-melted and annealed ternary FePdNi samples. Results for binary FePd ( = 0) are included for reference. Lines are drawn to guide the eye. Note that the annealed = 7.5 sample does not follow the trend line; the error bars for this sample are smaller than the symbol. ..................................................

........................................
........................................................ 103 calorimetry results at 20 K/min of ternary FePdNi samples in their as-arc-melted and annealed states. Samples are identified by their determined Ni content. DSC results for binary FePd ( = 0) are included for reference. For all annealed samples, consecutive heating scans are labeled scan” and “2 scan”. .............................................................................. 104Figure 44. X-ray diffractograms obtained for FePdCu samples in their a) as-arc-melted state, and b) annealed state. Samples are identified by their determined Cu content. Diffractograms obtained for binary as-arc-melted and annealed FePd = 0) are included for comparison, as well as reference patterns for fcc FePd ll as reference patterns for fcc FePd )0 FePd (JCPDS 03-065-9971 [48]), and A2 Fe (JCPDS 03-065-9971 [176]). ........................................................................ 106Figure 45. Variation with Cu content of unit cell volume, ratio, parameter, crystallite size () and microstrain () of the fcc (as-arc-melted) and L1(annealed) phases. Results for binary FePd ( = 0) have been included for comparison. Lines are drawn to guide the eye. ............................................. 107Figure 46. Magnetization as a function of applied field measured at room temperature for ternary FePdCu samples in their a) as-arc-melted and b) annealed states. Sample

s are identified by their determined Cu
s are identified by their determined Cu content. ) results obtained for binary as-arc-melted and annealed FePd ( = 0) are included for comparison. ................................................................................................... 108 with Cu content for as-arc-melted and annealed ternary FePdCu samples. Results for binary FePd ( = 0) are included for reference. Lines are drawn to guide the eye. ................................................................. 109Figure 48. Differential scanning calorimetry results at 20 K/min of ternary FePdCu samples in their as-arc-melted and annealed states. Samples are identified by their determined Cu content. DSC results for binary FePd ( = 0) are included for reference. For all annealed samples, consecutive heating scans are labeled “1st scan” and “2nd scan”. ............................................................................ 110Figure 49. Individual onset temperatures identified for the two endothermic features that conform the first-order peak observed in DSC for ternary FePdCu annealed samples. For the 5.8 annealed FePdCu samples, for both features is clearly identifiable, while for the = 7.9 FePdCu annealed sample deconvolution into two assymetric peaks was necessary to get an estimate of for the second event. ............................................................................. 111Figure 50. a) Fe-Pd-Ni ternary phase diagram at 82

3 K proposed by Horiuchi et alet alalong
3 K proposed by Horiuchi et alet alalong with the alloy compositions of the present study. b) Estimated compositions of the multiple phases present in the = 7.5 annealed sample: parent fcc phase (blue), product L1 phase (red), product new fcc phase (green). Horiuchi et al.'s L1 boundary is left for reference as a dashed line since it was used to estimate the compositions. ............................................ 114Figure 51. 773 K isothermal section proposed for the updated Fe-Pd-Ni ternary phase diagram ( represents the body centered cubic crystal structure while represents an fcc phase). L1 and fcc + L1 phase boundaries have been selected to match experimental observations, while phase boundaries ons are arbitrary. ...................................................... 115Figure 52. FePdCu ternary diagram with compositions from this study (annealed at 773 K) and those reported in literature for a similar annealing temperature (773 – 873 K) [107,109,180,181]. In gray text, the equilibrium phases expected for binary FePd are shown. The data points are color/symbol-coded to indicate the phases detected in these annealed alloys; blue (closed symbols) phase and green (open symbols) indicates an phase mixture. ............................................................................................... 116Figure 53. 773 K isothermal section proposed for the ternary Fe-Pd-Cu phase diagram (represents the bo

dy centered represents an fcc phase, r
dy centered represents an fcc phase, represents a B2 phase). The L1 phase boundary has been drawn to match experimental observations in this work and results reported by other authors [107,109,180,181]. Phase boundaries displayed for other regions are arbitrary.Figure 54. Evolution of the Curie temperature of L1 and fcc phases in a) FePdNi and b) FePdCu alloys as a function of Ni/Cu content. In (a), for the = 7.5 FePdNi annealed sample, note is plotted separately from the trend line presented for annealed L1 samples, due to the low volume fraction (~20%) of L1phase in this sample. ..................................................................................... 119Figure 55. Variation of the Curie temperature of a) fcc phase, and b) L1 phase, in near-equiatomic FePd-based alloys versus alloy composition, represented by the valence electrons per atom. The variation in the Curie temperature of fcc and phases in binary FePd as a function of composition, are included for comparison [26,182]. .................................................................................... 120Figure 56. Evolution of the onset temperature and enthalpy of the order-disorder fcc phase transformation in FePdNi as a function of Ni content. For the = 7.5 FePdNi annealed sample, note there is no data point for , as no measurable enthalpy associated with the L1fcc transformation was obtained .....................................

........................................
.................................................................... 123Figure 57. Evolution of the onset temperatures and enthalpy of transformation as a function of Cu content in FePd. For binary FePd, the transformation is fcc. For ternary FePdCu the transformation is + L1fcc. Lines are drawn to guide the eye. ................................................................................. 124Figure 58. a) X-ray diffractogram obtained for binary FePd annealed at 773 K for 100 h. (JCPDS 04-003-5130 [175]) and L1 FePd (JCPDS 03-065-9971 [48]) are included for comparison. b) and c) Detailed view of the regions 45° fits corresponding to L1 (blue) and fcc (orange) peaks, together with a cumulative fit (red). ...................................................................................... 130Figure 59. X-ray diffractogram obtained for binary FePd annealed at 773 K for 100 h and subsequently cold-rolled. A reference pattern for fcc FePd (JCPDS 04-003-5130 [175]) is included for comparison. ....................................................... 130Figure 60. Differential scanning calorimetry results at 20 K/min on FePd in the as-lled (orange, bottom) states. .................... 131 obtained for one of the undeformed FePd samples studied. Reference fcc FePd (JCPDS 04-003-5130 [175]) pattern is included for comparison. .......................................................................................

...... 132Figure 62. Isothermal DSC scan
...... 132Figure 62. Isothermal DSC scans at different temperatures for FePd samples. The DSC signal before the isothermal scan has been removed for simplification. ...... 133Figure 63. a) X-ray diffractograms obtained for FePd isothermally annealed in-situ in the DSC at 490 °C, 495 °C, 500 °C, 505 °C. Reference L1 FePd (JCPDS 03-065-9971 [48]) pattern is included for comparison. b) and c) Detailed view of the regions 45° indicative of the tetragonal symmetry, and no additional peaks that could signal the presence of an fcc phase. .............................................................. 134Figure 64. Variation with isothermal annealing temperature of lattice parameters, unit cell volume, and ratio of the L1 FePd phase. The unit cell volume of the starting fcc phase (undeformed, before isothermal treatment) has been included for reference. Lines are drawn to guide the eye. ............................ 135Figure 65. a) Isothermal fraction transformed, calculated from the area under the isothermal DSC peak for the transformation, for FePd. b) Linearized data following the JMAK model. ......................................................................... 136Figure 66. Arrhenius plot to determine the activation energy of the chemical ordering transformation in FePd from isothermal DSC data. ...................................... 137Figure 67. Isothermal DSC scans for different temperatures for cold-rolled

FePd samples. The DSC signal before the
FePd samples. The DSC signal before the isothermal scan has been removed for simplification. ............................................................................................... 138Figure 68. a) X-ray diffractograms obtained for cold-rolled FePd isothermally annealed in-situ in the DSC at 490 °C, FePd (JCPDS 03-065-9971 [48]) and Fe (JCPDS 00-019-0629 [73]) patterns are included for comparison. b) and c) Detailed view of the regions 45° and 67° tetragonal symmetry, and no additional peaks that could signal the presence of an fcc phase. .................................................................................................. 139Figure 69. Variation with isothermal annealing temperature of lattice parameters, unit cell volume, and ratio of the L1 FePd phase achieved in cold-rolled samples. The unit cell volume of the starting cold-rolled fcc phase (before isothermal annealing) has been included for reference. Lines are drawn to guide the eye.Figure 70. a) Isothermal fraction transformed, calculated from the area under the isothermal DSC peak for the transformation, for cold-rolled FePd. b) Linearized data following the JMAK model. ............................................... 141Figure 71. Arrhenius plot to determine the activation energy of the chemical ordering transformation in cold-rolled FePd from isothermal DSC data. ................... 142ffractograms for all compositions in their

as-milled and = 0.700292 Å radiation.
as-milled and = 0.700292 Å radiation. ........................ 159Figure 73. Variation of unit cell volume as a function of Fe content (at%). Reference values for fcc FeNi are included and connected by a linear fit (dashed line) [177]. Reference values for meteorite-derived L1 FeNi are also included for comparison [62]. ........................................................................................... 160diffractograms obtained with = 0.700292 Å radiation for a) milled samples, and b) annealed samples, showing the region near the (Bragg peak. An X-ray diffraction pattern for L1 FeNi, calculated from lattice parameters reported by Albertsen for meteoritic, near-equiatomic tetrataenite [62], is included for comparison. .................................................................. 161Figure 75. Synchrotron X-ray diffractograms obtained under anomalous diffraction conditions for the (FeNi)Ti annealed sample, with wavelength in the range = 1.74700 to = 1.87220 Å. a) region where superlattice ( peak region near fundamental (L10 peak. ........ 162Figure 76. a) Neutron diffraction pattern of cold-rolled annealed FeNi(Ti). Observed data points are shown with (bottom) and without (top) the calculated Rietveld fits overlaid. Differences between the obsed patterns are displayed in the cyan trace located below the peaks. Detailed views of the b) ) fundamental Bragg peak, and c) () fundamental Bragg peak,

derived from unprocessed FeNi(Ti) (red
derived from unprocessed FeNi(Ti) (red trace) and cold-rolled, annealed FeNi(Ti) (black trace), illustrating processing-induced broadening. ........................... 166Figure 77. Sample tetragonality as a function of Fe content, obtained from Rietveld refinements that employed an fct structural model to thdata of samples made from natural metal sources. Published data from natural tetrataenite are included [40,62]. The shaded region indicates a band of values and processed samples. ................................ 167Figure 78. Unit cell volume as a function of Fe content, obtained from Rietveld refinements that employed an fct structural model to thdata of samples made from natural metal sources. Published data from natural l well as for fcc FeNi [177]. ........................ 168Ni sample, and b) cold-rolled annealed FeNi sample. A semi-regular waveform background that both patterns. ....... 170Figure 80. Magnetization data obtained at 10 K for the cold-rolled FeNi(Ti) sample measured before and after annealing. Data were collected in the in-plane direction, both parallel and perpendicular to the rolling direction. ............... 171om the NWA6259 meteorite as a function of a) applied field, and b) temperature. In the as-received state, the meteorite is confirmed to be in the L1 state, while after two between room temperature and 700 °C, the sample is confirmed to be in the A1 state. There is an interme

diate magnetic state attained after the
diate magnetic state attained after the first heating/cooling cycle. Image adapted from [207] ........................................ 175Figure 82. Proposed sequence of phase transformation in Fe, where A1 is the parent cubic fcc phase, A6 is an intermediate tetragonal phase, and L1 is the product tetragonal chemically ordered structure ....................................................... 179Figure 83. Schematic representation of strain annealing procedure. Image adapted from[213]. ..................................................................................................... 191Figure 84. Instrumental setup at the X-16C beamline at NSLS (BNL). Image adapted from [221] ..................................................................................................... 218Figure 85. X-ray diffractograms of polycrystalline untextured powder Si standards for the calculation of the area correction factor. a) Small sample, with a size similar to samples of interest. b) Large sample, with an area sufficient to cover completely the incident X-ray beam. ............................................................ 227Figure 86. Plot of vs. for the Si standard. Data points were fit to an exponential function to determine the correction factor that should be applied to the integrated intensity of Bragg reflections at any given for samples smaller than X-ray illuminated area. ..................................

........................... 228Figure 87
........................... 228Figure 87. X-ray diffractogram for FeNi alloy cryomilled for different times. .............. 229Figure 88. Modified Williamson-Hall plot for FeNi cryomilled for different times. The broken lines represent the linear fit to the data, from which the intercept and slope are extracted to estimate the crystallite size and microstrain. ............. 230Figure 89. Evolution of crystallite size (left axis) and microstrain (right axis) as a function of cryomilling time for FeNi alloy. ................................................ 231LIST OF TABLES Table 1. Numerical values corresponding to conditions that determine the Avrami exponent. Adapted from [131]. ...................................... 54Table 2. Comparison of magnetic properties for L1 ferrous compounds and common rare-earth magnetic materials [10,76,77]. .......................................................... 62Table 3. List of samples studied in this dissertation. ........................................................ 66Table 4. Nominal and measured compositions (at%) of ternary FePdNi samples. .......... 97Table 5. Nominal and measured compositions (at%) of ternary FePdCu samples. ........ 105Table 6. Slope (Avrami exponent ) and intercept (ln ) of the linear fit to the isothermal data representing the chemical ordering transformation in FePd. The R value of the linear fits is provided. ...............................

........................................
.................................................. 136Table 7. Slope (Avrami exponent ) and intercept (ln ) of the linear fit to the isothermal data representing the chemical ordering transformation in cold-rolled FePd. The value of the linear fits is provided. .............................................................. 141ons of FeNi samples determined from laboratory X-ray = cubic lattice parameter, = coherently-diffracting crystallite = microstrain. ........................................................................................ 164Table 9. Rietveld refinement results for commercially available FeNi powder (Alfa Aesar), unprocessed FeNi and FeNi(Ti) samples, and processed FeNi and FeNi(Ti) samples. The goodness-of-fit and lattice parameters for each sample (fct) models. is the goodness of fit, R is the weighted-profile R-factor, R is the residual of least- and are lattice parameters, is unit cell volume, and is a metric of the tetragonality. ....................................... 166Table 10. Rietveld refinement results employing the tetragonal fct model for deformed pre-anneal and deformed post-anneal FeNi and FeNi(Ti) samples. is the goodness of fit, R is the weighted-profile R-factor, R is the residual of least-squares refinement, and are lattice parameters, is unit cell volume, and is a metric of the tetragonality. ...................................................................

..... 169) and Ge() [222]. .............
..... 169) and Ge() [222]. ................... 218Motivation The development of new materials to facilitate technological advancements relies on the understanding of structury correlations. This is particularly true in the field of permanent magnetic materials, in which reconciliation between the electronic/crystal/micro-structure of matter and the resultant magnetic properties has allowed enormous progress in the last 100 years [1], to produce stronger and lighter magnets to meet the ever evolving needs of society. Permanent magnets, used for energy conversion and generation, functionalize much of modern technology; they are important components in a wide variety of consumer items, medical equipment, and communication, military and alternative energy systems, among others [2–4]. The growing demand for advanced permanent magnets to power modern life, coupled with recent limited availability and price volatility of the rare-earth elements that constitute them [5,6], has resulted in an increased interest from the scientific ancommunities to realize novel and efficient magnetic materials made of earth-abundant elements [7,8]. The search for next-generation permanent magnets, certainly guided by knowledge gained by the magnetism community on structure-property relationships, considers two main requirements for candidate materials at the operating temperature: (the material should exhibit a strong and stable magneti

zation, and () there should be a source
zation, and () there should be a source of magnetic anisotropy to provide resistance to demagnetization. One avenue that conforms to these requirements is the use of ferromagnetic alloys and compounds that possess a low-symmetry crystal structure [7–9]. This dissertation examines fundamental aspects of ferromagnetic materials that advance the field of permanent magnetic materials Special attention has been devoted recently to near-equiatomic ferrous compounds with the tetragonal chemically ordered L1 crystal structure (Figure 1), such as L1 FePt, L1 FePd and L1 FeNi. Composed of alternating layers of the constituent elements stacked along the [ structure is chemically ordered to form a natural superlattice that exhibits a of magnetization along the tetragonal -axis. This appreciable magnetocrystalline anisotropy in L1-structured ferrous compounds, coupled to the large saturation magnetization donated by Fe, results in theoretical energy densities (a figure of merit to quantify the strength of a permanent magnet) that are comparable to those of current rare-earth-based magnets such as NdFeB or SmCo [10]. The prohibitively high cost and low availability of Pd and Pt limit the interest in L1-structured compounds made with these elements to applications such as magnetic recording and spintronic devices [9,11]. In the case of L1 FeNi, there is enormous interest for its use in bulk permanent magnet applications [7,12

,13], but there are still many challenge
,13], but there are still many challenges to overcome to achieve bulk synthesis of this material with scalable techniques and in laboratory timeframes. To date, only limited quantities of L1FeNi have been obtained by techniques such as neutron/electron irradiation [14–17] and monatomic layer deposition [18–22]. 3 Figure 1. Schematic representation of a) fcc (A1), and b) L1 crystal structures in binary AB alloys, where A and B are different elements. Half-shaded spheres represent an equal probability of lattice site occupancy by A or B atoms. For the L1 structure, planes along the -axis have been shaded to emphasize the chemical order. -type phases typically form by a chemical disorder-order phase transformation from a low-anisotropy face-centered cubic (fcc, Strukturbericht designation A1) parent phase (Figure 1) below a critical order-disorder temperature erature successful integration of L1-type ferrous materials into technological applications requires a comprehensive understanding of the A1 phase transformation. Tailoring the thermodynamic and kinetic aspects of this phase transition to meet laboratory (and subsequently industry) processing requirements is essential, with the accompanying need to understand the effect of any physical input used to drive the transformation on the structural characteristics and magnetic properties of the resultant L1 phase. To this end, this dissertation examines and quanti

fies the effects of intrinsic and extrin
fies the effects of intrinsic and extrinsic parameter variation on thermodynamic and kinetic aspects of the chemical order-disorder phase transformation in nominally equiatomic FePd and FeNi, as well as the structure-processing-property correlations associated with these systems. FePd is a good model system considering that the L1 phase can be achieved by traditional metallurgical processing techniques, such as isothermal annealing below the critical order-disorder temperature ( 660°C [26]). Therefore, FePd is employed in this dissertation to understand the effects of ternary alloying additions and of severe plastic deformation on the chemical order-disorder phase transformation in ferrous systems, to ultimately guide research efforts directed towards the achievement of the FeNi. Furthermore, the effects of ternary alloying additions and severe plastic deformation in FeNi-based alloys were also evaluated, to provide information regarding the kinetically limited chemical order-disorder transformation in this system. Overview of research The overall theme of this dissertation, understanding chemical ordering in FePd-based and FeNi-based alloys, was pursued by studying the effects of intrinsic and extrinsic modification on the structure, magnetism, and phase transformation character of FePd and FeNi samples at the near-equiatomic composition. These studies were categorized into three Aims, which will be explained

in the following paragraphs. of this
in the following paragraphs. of this dissertation examines and quantifies the effect of intrinsic ternary alloying on the A1 transformation of the model FePd system and on its magnetic properties. Bulk FePd alloys were modified with minor alloying additions of Ni or Cu (7 at%), and were characterized structurally and magnetically before and after annealing to induce the chemical ordering transformation ). Comparison of these modified samples with an unmodified binary FePd alloy allows relations to be drawn between amount and character of the ternary alloying additions and the structural and magnetic characteristics of the A1 and L1 phases. These characteristics include lattice parameters, long-range chemical order, saturation magnetization, magnetocrystalline anisotropy, and Curie temperature. Furthermore, an isothermal section (773 K) of the ternary Fe-Pd-Ni and Fe-Pd-Cu phase diagrams near composition was obtained from structural investigations performed on FePd-based alloys after annealing. The chemical disordering transformation (L1FePd-based alloys, studied through calorimetry, was determined to process with compositionally dependent order-disorder temperatures and enthalpies of transformation. of this dissertation examines the effect of extrinsic modification severe plastic deformation on the A1 transformation of the model FePd system. Bulk FePd samples in the L1 state were cold-rolled to deliver

a large amount of strain. They were subs
a large amount of strain. They were subsequently subjected to structural and calorimetric characterization to allow determination of the effect of plastic deformation on the long-range chemical order in FePd. Furthermore, studies of the evolution of chemical ordering (A1isothermal annealing of unmodified and cold-rolled FePd samples allowed determination of the different nucleation and growth mechanisms operating in undeformed and deformed samples, as well as quantification of the effective energy barrier for the transformation. Finally, of this dissertation investigates the effect of ternary alloying additions and severe plastic deformation in FeNi-based alloys. These experiments were carried out to provide fundamental information regarding the kinetically limited A1transformation in this system. Bulk FeNi-based alloys were processed through severe plastic deformation by cold-rolling or by cryomilling to deliver large amounts of strain to the material, and were subsequently annealed to induce the chemical ordering transformation. Advanced characterization techniques such as synchrotron X-ray e-of-flight neutron diffraction were used to probe the effect of severe plastic deformation and tewed identification of a new phase in the Fe-Ni binary system with tetragonal symmetry at the near-equiatomic composition, with formation triggered by deformation and annealing processing, allowing elucidation of possible mec

hanisms of the A1 chemical ordering tran
hanisms of the A1 chemical ordering transformation in the FeNi system. These results have important implications for the realization of bulk L1 FeNi in laboratory time-frames. Organization of the dissertation This document is divided into six Chapters. The Introduction (this Chapter), is Critical Literature Review (Chapter 2) which present pertinent background information regarding chemically ordered phases, with an emphasis on the L1 structure and on the chemical order-disorder phase transformation by which these phases form. Additionally, the FeNi and FePd material systems are introduced, and a review of previous work concerning L1-type chemical ordering in these material systems is provided. Furthermore, fundamental aspects of thermodynamics and kinetics of phase transformations and key concepts in magnetism are presented. Next, Chapter 3 -Experimental Techniques - provides a detailed description of the methods and equipment used in this dissertation for synthesis and processing of the samples, as well as for their structural, magnetic, and calorimetric characterization, including an overview of the terminology, operating principles, and analysis methods. A general description of the synthesis, processing and characterization parameters used in this dissertation is also provided in Chapter 3. Chapter 4 - Results and Discussion – is subdivided in three Sections, one for each Aim of this dissertation. The firs

t two Sections (Aims 1 and 2), present
t two Sections (Aims 1 and 2), present experimental data on the FePd system, pertaining to studies aimed at understanding the effects of ternary alloying additions and of plastic deformation on the structure, magnetic properties, and phase transformation character. The third Section (Aim 3) examines the effects of intrinsic and extrinsic modification to FeNi-based alloys to achieve the L1 phase. Each of the Sections in Chapter 4 starts with a detailed description of the specific synthesis and processing parameters applied to samples particular to that Aim, as well as presentation of the specifics of sample characterization. Structural, magnetic and calorimetric results and analysis of the results obtained for each Aim. Conclusions derived from the results of this dissertation are presented in Chapter 5 - Summary and Conclusion. Finally, recommendations for future work are discussed in Chapter 6. This dissertation focuses on understanding L1 phase formation in the FePd and FeNi-based systems, and identifying the factors that control the chemical ordering transformation by which the L1 phase forms. This current Chapter presents fundamental information relevant to the main topics of this dissertation, including a description of the crystal structure and of L1-chemical ordering transformations, an overview of the Iron-Palladium (Fe-Pd) and the Iron-Nickel (Fe-Ni) material systems, and a description of fundamenta

l aspects of phase transformations and o
l aspects of phase transformations and of magnetism. The Chapter has been divided in two main Sections. Section 2.1 -Literature Review- includes a general description of chemical order-disorder transformations, salient aspects of L1-type tics and properties of L1–structured FePd and FeNi, and an overview of the different factors reported to influence L1 chemical ordering in these material systems, including main results of prior literature that are pertinent for this dissertation. Section 2.2 includes an overview of fundamental thermodynamic and kinetic aspects of phase transformations, classification schemes of phase transitions, and a description of nucleation and growth processes. Additionally, this Section includes key concepts and terminology necessary for a better understanding of the magnetic behavior of ferromagnetic materials. Literature review Chemical ordering phase transformations are the main concern of this dissertation. In these transformations, the atoms in a crystalline alloy or compound, which are initially arranged randomly in the crystal lattice sites, rearrange at lower temperatures to produce a chemically ordered ific arrangement of atoms is also referred to as superlattice or superstructure. This current Section presents basic concepts associated with chemical order in binary alloys, introduces the Fe-Pd and Fe-Ni material systems used in this dissertation, and provides specifics of L1ch

emical order and the factors that affect
emical order and the factors that affect it. For clarification of key concepts related to thermodynamic and kinetic aspects of phase transformations, as well as terminology and notions used in the field of magnetism, the reader is directed to Section 2.2 Chemically order in binary alloys and the L1 crystal structure Binary substitutional AB solid solutions (where A and B are elemental species) which have a negative enthalpy of mixing favor the formation of unlike nearest neighbors [27], and thus chemically ordered phases form at low temperatures. In a chemically-ordered phase, atoms A and B are not randomly arranged in the crystal lattice but rather occupy specific positions. Some common chemically ordered structures are shown in oy systems in which they are found. 10 Figure 2. Common chemically ordered lattices: a) B2, b) L1, c) L1, d) D0, e) C0, e) C0. Chemical ordering may be accompanied by changes in the mechanical, magnetic, thermal, electric and optical properties of the material. Thus, observed changes in the elastic modulus (), hardness (), strength (), saturation magnetization (), Curie temperature (), magnetic susceptibility (), specific heat () and electrical resistivity ), among other characteristics, can be used to monitor a chemical ical Variations in these properties have also led to the prediction of chemically ordered structures in cases where structural investigations are inconclusive. For

instance, the L1superstructure in FeNi3,
instance, the L1superstructure in FeNi3, difficult to detect techniques due to the nearly identical scattering power of the constituent atoms, was predicted based on changes in resistivity and magnetic saturation upon cold-working annealed alloys [29]. In this dissertation, L1-type chemical order in the FePd and FeNi systems is crystal structure (space group ), which occurs in the vicinity of equiatomic compositions (50:50 at%) in selected systems, has a tetragonal lattice composed of crystallographic planes stacked along the [] direction which alternate in atomic species, Figure 3. Commonly represented in literature as a face-centered tetragonal (fct) structure, a body-centered tetragonal (bct) cell with half the volume of the fct (dashed lines in Figure 3) equally characterizes the L1 structure. By crystallographic convention, the smallest primitive unit cell should be taken to represent a crystal lattice. phases however, it has been common practice in the literature to adopt the fct cell, that possesses the same orientation of the basis vectors as a face centered cubic (fcc) crystal structure (space group , Strukturbericht designation A1). In this manner, an easy comparison can be made to the A1 phase, which is usually the parent phase. In this dissertation, the face-centered tetragonal representation of the L1Figure 3. Schematic representation of a) fcc (A1), and b) L1 crystal structures, where A and B

are constituent elements. Half-shaded s
are constituent elements. Half-shaded spheres represent an equal probability of lattice site occupastructure is shown with the same orientation of the basis vectors as that of the cubic unit cell (fct representation), and dashed lines are used to delineate the primitive Generally, the L1 phase forms by a nucleation and growth process from a parent A1 phase below a critical chemical order-disorder temperature (erature (formation of the L1 phase, a decrease in the number of equivalent positions in the crystal lattice reduces both the translational and the point group symmetry, resulting in crystallographic translational and orientational structural domains [24]. The first case (translational domains) refers to out-of-step domains sepaboundaries (APB), which occur when two L1 nuclei configured in a different sub-lattice grow and coalesce (Figure 4a) [24,28]. The latter case (orientational domains) relates to ons of the tetragonal -axis of the L1 phase along any of the three � of the parent cubic cell. Therefore three crystallographic “variants” are possible (Figure 4b), that generate orientational domains [23,24]. Translational and orientational domains have an important effect in the development of the microstructure, and therefore on the magnetic and mechanical properties of the L1 compound. s found upon formation of an L1 phase. a) Translational domains, separated by an anti-phase boundary. b) Orientat

ional domains, for which the tetragonal
ional domains, for which the tetragonal axis is directed along any of the three different basis vector of the cubic cell. These domains are Overview of the Iron-Palladium and Iron-Nickel material systems This Section presents a description of the Fe-Pd and Fe-Ni material systems studied in this dissertation. Of special interest, the L1 phase in near-equiatomic FePd and FeNi alloys will be discussed. L1 FeNi is of great interest for advanced permanent magnet applications due to its compelling magnetic properties together with the low cost and good availability of the constituent elements. L1 FePd also possesses interesting magnetic properties, yet its use in technological applications is limited due to the high cost of Pd. From the fundamental science point of view however, FePd is an excellent model system to study chemical order-disorder transformations. Compositional and temperature stability ranges of the L1 phase in FePd and FeNi are presented here, as well as structural and magnetic changes in near-equiatomic FePd and FeNi alloys that arise as chemical ordering. The Iron-Palladium (Fe-Pd) system The Fe-Pd phase diagram (Figure 5) presents five stable equilibrium solid phases, namely ). The phases A2) with a limited solid solubility of Pd (3.7 at%). The –(Fe, Pd) phase is a face-centered-cubic phase that is stable for all compositions at high temperatures. At low temperatures, two chemically -(FePd) with

the tetragonal structure and -(FePd) wi
the tetragonal structure and -(FePd) with the cubic L1 structure (space group ). The congruent L1 and L1 ordering occurs at off-stoichiometric compositions (~59 at% Pd ) at 790 °C and 820 °C, respectively. Both chemically ordered phases have large homogeneity ranges, between 50 and 60 at% Pd for the L1Figure 5. Binary Iron-Palladium phh. L10 phase formation in bulk FePd samples has typically been accomplished through isothermal annealing in the temperature range 400-600 °C for periods ranging from 1 to 200 h [10,33–38]. Time-transformation-temperature () curves available for the near-equiatomic FePd composition (Figure 6) indicate that the maximum rate of transformation occurs near 525 °C, taking ~1 h to achieve a 90% volume fraction of the phase [39]. The transformation into the L1 phase of bulk near-equiatomic FePd has also been accomplished by cooling thorough the order-disorder transformation temperature to room temperature at slow rates (0.5-2 K/min), with higher degrees of L1chemical order obtained for the slower cooled samples [40]. 15 curves for near-equiatomic FePd. L1 volume fractions of 5, 20, 50 and 90% are presented as a function of temperature (Celsius) and time (h) of isothermal annealing. Adapted from g. Adapted from . The L10 chemical ordering transformation in near-equiatomic FePd alloys that occurs upon isothermal annealing, for undercoolings ( = anneal) as large as 250 K, has been re

ported to proceed a nucleation and grow
ported to proceed a nucleation and growth mechanism [10,23,33,34,41,42]. In this process, both parent A1 and product L1 phases are coherent along boundaries. The coherent nucleation of the tetragonal L1 phase within the cubic A1 matrix results in lattice strains that are initially minimized by attainment of a non-equilibrium ratio in the L1 phase, and that at later stages of the transformation are relaxed by the formation of a self-accommodating array of twins [23,37,41–44], as will be seen in greater detail in Section 2.1.3.3. Vlasova et al. have reported on L1 phase formation in single crystal [45,46] and polycrystalline severely deformed FePd samples [46] by study of samples subjected to isothermal annealing at 450-600 °C. The results of these studies suggest that the A1 transformation in FePd might actually occur a two-stage process, with the existence of an intermediate tetragonal, but chemically disordered phase (space group I4/mmm, Strukturbericht designation A6) that bridges the A1 and L1 structures, Figure 7. The mechanism of phase transformation proposed by these authors involves an initial A1A6 cooperative displacement phenomena (diffusionless) followed by a chemical ordering transformation process A6Figure 7. Schematic representation of the A1 transformation in FePd through ugh . Regardless of the mechanism of chemical ordering, the A1 transformation in near-equiatomic FePd has been reported to

produce a tetragonal distortion as rati
produce a tetragonal distortion as ratio of ~0.96 [47,48]. Furthermore, a change in unit cell volume is also reported upon formation of the L1 phase; a relative increase in the unit cell volume of 0.55-1.54% has been reported for a Fe composition by Hultgren [48] and Raub [49], while a relative decrease of 0.23% has been reported for the Fe composition by Okamoto [26]. These structural changes are accompanied by changes in the magnetic properties. The reduced symmetry of the tetragonal L1 structure provides a strong uniaxial magnetic character, with the tetragonal -axis becoming an easy-direction of magnetization. Thus, upon L1 chemical ordering, FePd experiences an increase in magnetocrystalline anisotropy of two orders of magnitude, achieving values between erg/cc [10,45,50,51]. This high magnetocrystalline anisotropy donates a high resistance to demagnetization, essential for permanent magnet applications. Changes in the Curie temperature of FePd alloys upon L1 chemical ordering have also been reported; however, it is controversial whether increases or decreases during the transformation. Wang ng e temperature for the A1 phase in comparison to that of the L1 phase in bulk Fesamples, while Vlasova et al. [44] show a higher Curie temperature for the L1 phase in equiatomic FePd single crystals. Both papers agree that the Curie temperature values previously reported in the literature for A1 and L1 phases are

highly conflicting, and [45,46] propos
highly conflicting, and [45,46] propose that this may be due to the existence of the intermediate The Iron-Nickel (Fe-Ni) system The study of phase equilibria in the Fe-Ni system below 400 °C is severely hampered by the slow diffusion of Ni in Fe below this temperature [52,53]. The existence of a low-temperature superlattice was first suggested by Dahl in 1936 [29], drastic changes in resistivity for different processing conditions that mirrored the behavior of the AuCu system, known to exhibit low-temperature superstructures. These changes were observed for samples in the compositional range 40-90 at% Ni, but were most drastic at the Fe composition. From these results, an L1-type superlattice was hypothesized, without discarding the prospect of a near-equiatomic chemically ordered-structure. Later, in a thorough study involving structural, magnetic and resistivity measurements, Wakelin and Yates were able to confirm the existence of the L1superlattice in the compositional range 65 to 81 at% Ni, and a tentative suggestion was made about an ordered structure of a different kind at the 50 at% Ni composition (and even 45 at% Ni) [54]. It was not until the 1960’s that unambiguous evidence of the existence of an L1-type structure in FeNi was achieved from the study of samples irradiated with neutrons/electrons in a magnetic field at low temperatures eratures rally occurring L1 FeNi was later detected in a variety

of stony, iron, and stony-iron meteorite
of stony, iron, and stony-iron meteorites [57–59] subjected to cooling rates as slow as ~0.1– year [60,61]. The slow cooling over timescales of millions of years [60,62,63] was reported to foster the development of tetragonality and of L1-type long-range chemical order from a chemically-dically-dInvestigations on meteoritic and irradiated FeNi samples allowed Yang les allowed Yang to propose in 1996 the modern Fe-Ni phase diagram that includes phase equilibria information below 400 °C, Figure 8. In this phase diagram, an -Fe phase with the body-centered-cubic crystal structure shows a limited solid solubility of Ni ( 5.8 wt%), while -(Fe, Ni) phase with the face-centered-cubic structure is stable for all compositions at high temperatures. A tricritical point is found around 462 °C and 48.8 wt% Ni, below which the -phase splits into low-Ni paramagnetic () and a high-Ni ferromagnetic (phases. This phenomenon has been associated with magnetic contributions to the Gibbs free energy of the phase [53]. Congruent chemical ordering occurs very close to the stoichiometry below a critical ordering temperature of 516 °C, producing an L1type ( FeNi phase () - also known as tetrataenite - is presented in the modern Fe-Ni phase diagram as a metastable line compound existing on the high Ni content boundary (~52 wt% Ni) of the miscibility gap () below a critical ordering temperature of 320 °C [52]. Recent calorimetric studie

s of the L1A1 chemical disordering trans
s of the L1A1 chemical disordering transformation in meteoritic samples [64] indicate that the L1 phase is enthalpically stabilized relative to the A1 phase, and is at the edge of enthalpic stability relative to a mixture of -Fe and L1 FeNi. However, the entropy contribution to the Figure 8. Fe-Ni low temperature phase diagram with a bcc phase (), fcc phase ( ordered FeNi) phase and L1 ordered FeNi ( ordered FeNi (. In laboratory timescales, the A1 transformation in FeNi is highly kinetically limited due, in part, to the low order disorder temperature of 320 °C [62] together with the extremely sluggish diffusion at that temperature [16,52]. From diffusivity data, it has been estimated that at 300 °C it takes 2600 years for one atomic jump to occur in to occur in fabrication of limited quantities of L1been substantiated in materials subjected to the non-scalable processing techniques of [14,15,17,55,56] or by monatomito produce ultra-thin films [18–21,51]. L1 phase formation in irradiated samples was attributed to enhancement in the rate of diffusion fostered by a high density of extrinsic lattice defects [17,56,65]. Since chemical ordering processes require local atomic rearrangements, the presence of vacancies that allow atomic motion [27,66,67] can be expected to facilitate L1 chemical ordering. In factagreement between the results of their studies on the rate of chemical ordering in electron-irradiated F

eNi samples and those resulting from app
eNi samples and those resulting from application of a model of long-range chemical ordering that is proportional to the vacancy concentration [68]. Recent studies on the L1A1 transformation in meteorite-derived L1 FeNi have also concluded that the transformation rate in this system is indeed limited by the creation and migration of vacancies [64].Thus the application of processing techniques that result in the formation of excess vacancies, are anticipated to have a positive impact on the rate of chemical ordering in FeNi. However, it has been proposed that magnetic interactions in this system hinder the formation of vacancies; Yang and Goldstein [69] performed theoretical analyses of the contribution of magnetism to diffusion in fcc FeNi alloys, based on experimental physical, mechanical and magnetic properties (elastic constants, lattice parameters and magnetization). Their analyses reveal that the magnetic contribution to the free energy of vacancy formation and migration, and thus to diffusivity, varies considerably with temperature and composition (Figure 9). Their calculations suggest that vacancy formation is less favorable in the ferromagnetic state in comparison to the paramagnetic state, while vacancy is reported to be more favorable below the Curie temperature . For near-equiatomic FeNi (50:50 at%) alloys and those with higher Ni content, the net contribution of magnetism to the free energy for diff

usion is positive (Figure 10), which res
usion is positive (Figure 10), which results in a significant reduction in the diffusivity . These results emphasize the difficulties associated with the L1-type chemical ordering process in FeNi, and demonstrate that efforts towards the bulk synthesis of the phase must identify suitable processing routes to generate excess vacancies that can be effectively used to facilitate atom rearrangement for long-range chemical order. Figure 9. Magnetic contribution to the free energy of a) vacancy formation, and b) vacancy migration, in fcc FeNi alloys as a function of temperature and composition. The Curie temperature () is the critical temperature at which there is a ferromagnetic to paramagnetic transition. Image adapted Figure 10. Magnetic contribution to the free energy for diffusion in fcc FeCurie temperature () is the critical temperature at which there is a ferromagnetic to paramagnetic transition. Image adapted from age adapted from . In recent years, several studies have experimented with different processing techniques that enhance diffusion in pursuit of the L1 phase in bulk near-equiatomic et al. [70] utilized room-temperature mechanical milling to induce excess vacancies that could favor the formation of L1 FeNi. The L1 phase was not detected through X-ray diffraction, and a positron annitechnique revealed that the mechanical milling process did not effectively generate a high vacancy concentration. O

n the other hand, Lee et al. [71] applie
n the other hand, Lee et al. [71] applied high pressure torsion to a mixture of Fe-Ni powders, as well as to bulk FeNi alloys, to deliver intense strain that could favor atomic diffusion. These authors incorporated cobalt in some of their samples, considering it is a major impurity element found in some L1-containing meteorites and it may play an important role in L1 phase formation. They report achievement of the L1FeNi phase as evidenced from laboratory X-ray diffraction and transmission electron diffraction results, but it remains controversial whether the detected L1-like reflections are a signature of the L1 FeNi phase or if they correspond to easily formed oxides in this system, confirmed to exhibit Bragg reflections that overlap with those examined by the authors [72–74]. Furthermore, Makino et al. [75] have pursued synthesis of the L1 FeNi phase starting from an amorphous Fe41.30.7 alloy, considering the high atomic diffusivity possible upon recrystallization from an amorphous state. The authors report attainment of 8-13% volume fraction of the L1 FeNi phase, based on structural on diffraction, and magnetic characterization by magnetometry. These results have originated much debate, as the different structural signatures that the authors recognize to originate from the L1 phase either overlap those of easily formed oxides, or are much stronger than those expected for FeNi with perfect chemical order.

Results available on meteoritic and irra
Results available on meteoritic and irradiated L1 FeNi demonstrate that L1chemical ordering in this system is accompanied by a tetragonal distortion of = 1.0036, and a reduction in the unit cell volume of ~0.5% in comparison with the A1 counterpart [62]. Concomitant with the L1 chemical ordering process, an increase in magnetocrystalline anisotropy is reported, with values of 1 - 1.3 x 10 erg/cc reported for phase [15,76]. Additionally, changes in also occur upon formation of the L1FeNi phase; calorimetric investigations of meteoritic L1 Fe samples indicate a Curie temperature of ~532- 557 °C [76,77], while the Curie temperature of an A1 phase of the same composition is ~452°C [78]. It should be noted however that the experimentally-identified Curie temperature of L1 FeNi is a “kinetic” Curie temperature, concurrent to a kinetically-limited L1A1 disordering process. Theoretical calculations have estimated that of L1e range 640-730 °C [76,77,79]. -type chemical order In this Section, basic structural, microstructural, and magnetic aspects of L1chemical ordered phases will be covered. The discussion includes an explanation of the degree of order, structural signatures of L1-type chemical order and the challenges associated with the detection of L1-FeNi, the evolution of microstructure during the transformation, and magnetic domain structures in L1 materials. Degree of chemical order As a chemically-ordered

atomic arrangement, where atoms are mean
atomic arrangement, where atoms are meant to occupy -sites for A atoms and -sites for B atoms), the L1 phase may be characterized by the long-range order parameter, which quantifies the degree of chemical order. Considering the atomic fractions of A and B species as and , the concentration of and -sites as and and sites occupied by the right atom (A or B, respectively) as parameter is defined as [47]: BAxrxrrrLRO ( 1 ) parameter ranges from a value of zero (no chemical order) to a value of unity (perfect chemical order). The factors that affect the parameter are composition and temperature [28]. Considering the L1 phase, perfect order can only be achieved for the equiatomic composition, with equal atomic fractions of A and B species guaranteed. Off-stoichiometry can be accommodated by anti-site disorder or vacancies, decreasing parameter. Additionally, perfect may only be realized at absolute zero temperature [27,47]. Increasing temperature results in increased thermal vibrations of the atoms around their equilibrium positions that may result in atom interchange and an overall reduction in . The way in which the parameter decreases with increased temperature provides insight into the thermodynamic character of the chemical order-disorder transformation (See Section 2.2.1 for details on 1-order phase transformations). In some cases, the chemical order-disorder transformation has a 2order character,

with the parameter decreasing continuou
with the parameter decreasing continuously up to the order-disorder transition temperature (for example, see CuZn in Figure 11). In other cases, the chemical order-disorder transformation has a 1-order character, and the parameter experiences a modest decrease with temperature until reaching , where it abruptly drops to zero (for example, see AuCu25 Figure 11. Variation of parameter with temperature for two different chemically ordered alloys. Image adapted from . 2.1.3.2 Detection of L10-type chemical order most direct method usually used to characterize L1type chemical order. A description of different scattering techniques is provided in Experimental Techniques-. Two very distinct features appear in a diffraction pattern when a material system undergoes the A1 transformation. The first feature, associated with chemical order itself, is the appearance of “superlattice” or “superstructure” peaks. These are Bragg reflections that were forbidden in the A1 structure due to the randomness of atom occupation on the fcc lattice sites, but that are no longer forbidden in structures where the A atoms occupy the -sites and the B atoms -sites [28,81]. The intensity of these superstructure lines is proportional to the absolute value of the difference between the atomic scattering factors of the A and B elements, as explained in greater detail in Appendix 2. The second feature detected in a diffraction pattern upo

n L1-type chemical ordering is the split
n L1-type chemical ordering is the splitting of some fundamental Bragg reflections that relates to the tetragonality of the L1 structure, for which the reduced symmetry decreases the number of equivalent crystallographic planes. The magnitude of the peak-splitting is proportional to the axial ratio of the L1 crystal structure. An example of X-ray diffractograms expected for polycrystalline A1 and L1structured FePd, showing the appearance of superlattice peaks an chemical ordering, is presented in Figure 12. Figure 12. Calculated X-ray diffraction patterns using the FullProf Suite patterns using the FullProf Suite for: a) A1 FePd, and b) L1 FePd. Asterisks designate superlattice peaks, while arrows show peak-splitting. Microstructure in L1 chemically ordered alloys Typical microstructures transformation a nucleation and growth mechanism are presented in Figure 13. A cross-hatched contrast referred to as “tweed” contrast (Figure 13a) is observed at the early stages of the transformation. At this stage, all three variants of the L1 phase form, but as the transformation proceeds two of them dominate [41]. Later stages of the transformation are characterized by micro (nm-scale) and macro-twins (m-scale) which arise from the coalescence of discrete ordered regions with different orientation of the -axis, producing what is known as a “poly-twinned” microstructure (Figure 13b and c). Usually there is a high densi

ty of anti-phase boundaries within the t
ty of anti-phase boundaries within the twin plates of these structures (Figure 27 b) c) es resultant from the A1 transformation. a) Tweed contrast, b) Micro-twins and anti-phase boundaries, c) Poly-twinned microstructure. Images adapted from inned microstructure. Images adapted from . The observed microstructures in L1-forming alloys can be explained in the context of a nucleation and growth process of a tetragonal phase within a cubic matrix, with the important property of coherency of the chemical ordering process in FePd and FePt bulk samples through data obtained from X-ray diffraction (XRD) and transmission electron microscopy (TEM) [23,41,43] has allowed identification of the following sequence for the evolution of the microstructure during the A1 transformation. At the early stages of ordering, the -ordered tetragonal phase with a ratio different from the equilibrium value nucleates in the form of small plates along {} planes of the parent A1 matrix. A non-equilibrium ratio may occur in order to decrease the high activation energy barrier associated with the nucleation of the tetragonal phase. The with the nucleation of the L1 phase give rise to the observed tweed contrast. Vlasova [46] have argued that the tweed contrast at the early stages of ordering is actually related to a pre-transitional tetragonal but chemically disordered A6 phase, as discussed in Section 2.1.2.1. As the ordering trans

formation proceeds, two L1 variants domi
formation proceeds, two L1 variants dominate due to stress-affected preferential growth and are arranged in alternating bands along the } planes, Figure 14. Growing L1-ordered particles impinge and coalesce within the bands, producing a high density of anti-phasconjoin along coherent twin boundaries to form the poly-twinned structure (Figure 13b). Within a particular grain, twin clusters with different combinations of the variants impinge at macro-twin boundaries (Figure 13c). Upon L1 growth, the ratio will gradually shift towards equilibrium with changes persisting even after the A1 phase has Figure 14. Schematic representation that shows mutually orthogonal nuclei of L1al nuclei of L1. 2.1.3.4 Magnetic domain structure in L1ostructure commonly found in bulk polycrystalline L1 ferrous samples (described in the previous Section) produces a unique magnetic domain structure which has a direct effect on the mechanism of magnetization reversal and hysteresis behavior [41,42] (see Section 2.2.3 for key magnetism concepts and theoretical background). This domain structure is composed of modulated magnetic regions in which the easy axis of magnetization varies quasi-periodically on a scale of 10 to 100 nm [23,41], Figure 15. Two types of interfaces between magnetic domains, called “domain walls” exist in this microstructure; immobile or “frozen” domain walls, associated with the {} micro and macro-twin boundaries,

and serrated mobile domain walls which
and serrated mobile domain walls which can traverse the micro-twin plates. The frozen domain walls are those between magnetic domains with magnetization vectors at 90° or 180° with respect to each other. Mobile domain walls that are located within the micro-twin plates are 180° walls and are expected to control magnetization reversal. Pinning of these mobile walls at APB may control the coercivity. The thickness of the domain walls for L1 FePd has been estimated at 5 nm [42]. Figure 15. Schematic representation of the magnetic domain configuration in L1poly-twinned alloys. Thin lines represent micro-domain boundaries, thick lines are macro-domain boundaries and serrated lines are mobile domain-walls cutting across the array of micro-twins. Arrows represent magnetization vectors. Image adapted from ation vectors. Image adapted from 2.1.3.5 Challenges associated with-type chemical order As presented in Section 2.1.3.2, detection of L1 phases is usually accomplished using scattering techniques that can provide direct evidence of the presence of long-range chemical order and of tetragonal symmetry through observation of superlattice reflections and peak-splitting in a diffraction pattern. However in some alloys, such as FeNi, direct identification of the L1 phase through diffraction methods is extremely challenging considering the subtleness of both of these features. The small difference in X-ray scattering

factors and in neutron scattering lengt
factors and in neutron scattering lengths of Fe and Ni results in superlattice peak intensities that are usually below the detectability limit in most instruments. Calculations considering a sample consisting of 100% volume fraction of untextured (isotropic) L1 FeNi phase with perfect chemical order ( = 1) indicate that the ratio that of the main fundamental Bragg reflection is very small, with values of 0.30% for X-ray diffraction and 0.28% for neutron diffraction. Similarly, the very slight tetragonality of the L1 FeNi phase = 1.0036 [62]) results in subtle peak broadening rather than peak-splitting, undetectable in most cases unless a very high instrumental resolution is guaranteed. Figure 16 presents calculated X-ray and neutron diffraction patterns for L1 FeNi, which demonstrate the challenges mentioned above. Detection of the L1 FeNi phase by diffraction techniques is expected to be even more arduous in samples that contain small fractions of the phase, or in specimens in which the phase has a low degree of long-range chemical order. This Section presents different characterization methods reported in the cal detection of the L1 FeNi phase, together with experimental conditions that are anticipated to facilitate the unambiguous identification of chemical 31 Figure 16. Simulated a) X-ray and b) neutron diffraction patterns for L1 FeNi considering lattice parameters reported by Albertsen orted by Alberts

en . Superstructure peaks (marked by ast
en . Superstructure peaks (marked by asterisks) have intensities near the signal-to-noise limit, and there is no evident peak-splitting, except at very high angles in X-ray diffraction (e. g. () peak). The X-ray diffraction pattern was calculated using the FullProf Suite ed using the FullProf Suite , while the neutron diffraction pattern was simulated by collaborators at Rutherford Appleton Diffraction techniques Special circumstances, in terms of sample characteristics or of experimental favor the detection of L1 FeNi using diffraction techniques. For instance, the use of a single-crystal L1 FeNi sample measured with its axis parallel to the diffraction vector may facilitate the detection of superlattice Bragg peaks. Indeed, superlattice peak observation has been reported for X-ray diffraction [14,83] and selected area electron diffraction (SAED) [15] experiments performed on neutron-irradiated FeNi single-crystal specimens, confirming that these samples exhibited -type chemically ordered structure without an appreciable tetragonality. Meteoritic-derived single-crystal FeNi lamella probed by XRD [57,84–87] and by SAED [88] have also demonstrated the presence of superlattice peaks, s-type chemical ordering in these samples. The use of synchrotron X-ray diffraction and neutron facilities, at which higher fluxes, better -spacing resolutions, and improvecomparison with those characterizing laboratory-based diff

raction methods (see Section 3.3.2 for m
raction methods (see Section 3.3.2 for more information), may facilitate the detection of the L1 FeNi phase. This assertion is especially true if: ) a single-crystal specimen is not available, ) the sample contains only a small volume fractiodegree of long-range chemical order. An additional advantage of synchrotron X-ray probes is that they feature energy tunability, allowing the selection of a suitable wavelength to manipulate the atomic scattering factor for easier detection of superlattice peaks originating from an L1 FeNi phase. This manipulation makes use of so-called “anomalous” or “resonant” diffraction conditions which occur near the absorption edge of an element, at which its scattering factor is greatly decreased (further details provided in Appendix 2). Thus, selection of the incident X-ray with an energy that is close to the absorption edge of Fe or Ni may increase the difference in the X-ray scattering power of these two elements, resulting in enhanced superlattice peak intensity. In the case of neutron diffraction, for which different elemental isotopes may scatter neutrons differently (as will be seen in greater detail in Section 3.3.2), manipulation of the atomic scattering factor may be accomplished via selection of a suitable isotope to form the FeNi alloy. For example, natural Ni has a positive neutron ) while that of Ni is large and negative (Ni = 10.3 fm, 8.7 fm [89]). Therefore, Ni can b

e used together with natural Fe (Fe = 9.
e used together with natural Fe (Fe = 9.45 fm) to Ni alloys with a geometric structure factor that provides a very high sensitivity to the long-range chemical order. To visualize this phenomenon, Figure 17 presents a neutron diffraction pattern calculated for isotropic, perfectly ordered L1Ni. It can be seen that this pattern is dominated by superstructure peaks, with the main superlattice () peak exhibiting an intensity that is ~ 890 times that of the main fundamental Bragg reflection. This strategy allows unequivocal determination of the presence of L1 chemical order even for samples with small L1 phase volume fraction or with imperfect long-range chemical order. Figure 17. Neutron diffraction pattern for L1 FeNi, simulated by collaborators at Rutherford Appleton Laboratory, considering lattice parameters reported by Albertsen by Albertsen . In contrast to the diffraction pattern for L1 FeNi made from natural sources (Figure 16), the diffraction pattern for L1 FeNi is dominated by superlattice peaks, and the intensity of the fundamental Bragg reflections is greatly decreased due to the near null-scattering Mössbauer spectroscopy The planetary science community has made extensive use of Mössbauer spectroscopy for the detection of L1 FeNi in selected meteorites. This technique is based on the Mössbauer effect which utilizes the recoil-free resonant absorption and emission of gamma rays in solids. Mössbauer spe

ctroscopy probes changes in the energy l
ctroscopy probes changes in the energy levels of an atomic nucleus in response to its environment, with an extremely high degree of accuracy (up to 13-15 decimal places) [90]. Thus, this technique can provide information on the local electronic, magnetic, and structural properties within the material, including chemical ordering phenomena. In a Mössbauer spectroscopy experiment, a gamma-ray source is oscillated to modulate the energy of the emitted ray in small increments. The emitted rays are directed to an absorber (sample), which must be composed of the same isotope as the emitting source. A detector measures the intensity of the beam transmitted through the sample. When the energy of the emitted gamma-rays matches the energy of a nuclear transition in the absorber, the gamma-rays are absorbed, resulting in a peak in the absorption spectra collected [90,91]. The absorption spectra is analyzed and compared to standard spectra characteristic of different phases or compounds available in databases, to determine the chemical environment of the resonant atom. Thusanalysis can be accomplished, as well as determination of the concentration of the resonant element in different phases [91]. Indeed, several authors have employed the Mössbauer spectroscopy technique for the characterization of FeNi-based meteoritic samples, demonstrating the presence of a phase in which all Fe atoms have identical nearest-neighbor confi

gurations (i.e. a phase with a chemicall
gurations (i.e. a phase with a chemically ordered structure) [57,59,62,86,87,92]. Coupled with microprobe analysis, which confirmed that this ordered phase had a near-equiatomic FeNi composition, the presence of the L1FeNi phase in these meteorites was unambiguously demonstrated. The high resolution of this technique certainly allows for the detection of small volume fractions of the L1 FeNi phase in processed samples, even when they have imperfect long-range chemical order. As mentioned in Section 2.1.1, chemical ordering phenomena may result in changes in the electrical, mechanical, magnetic, optical and thermal properties of an alloy. These changes can be monitored in processed FeNi samples to indirectly detect the formation of an L1-type chemically ordered phase, but they do not provide unambiguous proof of its existence. However, considering the difficulties associated with the direct detection of L1 FeNi, indirect characterization methods initially provided a means to indicate the existence of the L1 phase in the Fe-Ni system (as explained in Section 2.1.2.2). Some techniques that have been employed in the literature for indirect detection of the L1 FeNi phase in meteorites and in irradiated samples are presented below. Current efforts directed towards the laboratory synthesis of bulk L1 FeNi could make use of these techniques to provide an indication of the achievement of chemical order, as long as a hi

gh volume fraction of the phase is prese
gh volume fraction of the phase is present in the processed samples. In the case of electrical and magnetic measurements, a high degree of long-range chemical order is also necessary for detection. Furthermore, in magnetic measurements, a proper microstructure as well as a preferential direction for the alignment of the -axis of the L1Reflected light polarized microscopy The observation of optical anisotropy of the L1 FeNi phase has been used in the meteoritic literature to differentiate it from the kamacite (bcc -Fe with up to 7 at% NiNib. Electrical resistivity measurements A decrease in the electrical resistivity of an Fe alloy with increased neutron irradiation dose was observed in experiments performed at different temperatures below 321 °C [96,97]. This phenomenon, not observed in alloys irradiated at higher temperatures, was attributed to the establishment of L1-type long-range order as a result of the irradiation process, and from these experiments the critical order-disorder temperature of FeNi was confirmed to be near 320 °C. Magnetic property measurements The magnetic properties of near-equiatomic FeNi are reported to be affected by the development of L1-type chemical order. In fact, it was the analysis of the evolution of magnetic anisotropy energy of an Fe single crystal as a function of neutron irradiation dose that allowed Néel, Paulevé and collaborators in the 1960’s to suggest the achieveme

nt of a superstructure by neutron/electr
nt of a superstructure by neutron/electron bombardment [17,55,56,65]. In these reports, observation of the development of a strong uniaxial magnetic anisotropy in samples irradiated below a critical temperature of 320 °C indicated the achievement of a low-symmetry phase, confirmed to be of an L1-type by direct characterization techniques. Development of the L1 phase was observed to be accompanied by an increase in the coercivity, from a value of 0.5 Oe for non-irradiated samples to values of 40 Oe for irradiated FeNi [65]. Further, studies of the magnetic properties of meteorites FeNi phase have shown a change in the saturation magnetization and Curie temperature when undergoing the L1A1 transformation. The chemically ordered state is reported to exhibit lower (~15% measured at 5 K) and (~20-34%) and room-temperature (1100 – 4000 Oe versus essentially zero) relative to the chemically disordered state [76,77,98,99]. Factors that influence the A1 phase transformation in ferrous alloys This Section describes changes observed in the A1 phase transformation in ferrous systems, particularly in FePd, as a response to compositional modifications and external stimuli, such as pressure and magnetic field. These observations provide a basis for understanding how the A1 transformation can be tailored, and give insight onto suitable conditions required for the achievement of the L1The effect of compositional modificati

on by substitutional ternary transforma
on by substitutional ternary transformation Intrinsic modification of alloys has been used for millenia to tailor technical properties of interest. In the case of ferrous L1-formingalloys, intrinsic modification has been commonly employed to either tune the magnetic properties, or to modify the thermodynamics and kinetics of the chemical order-disorder transformation to favor processing of the L1 phase. For instance, much research in the FePt system has considered the use of ternary additions to tailor to meet the requirements of the hard-disk drive industry, without compromising its hard magnetic properties. Additions of Au and Ag to FePt have been reported to reduce [100,101]. It was suggested that Au and Ag have a high mobility, and that when incorporated into FePt they tend to diffuse out to the grain boundaries, leading to enhanced atomic migration. Addition of Cu to FePt was reported to increase the coercivity and accelerate the L1 ordering process [100,102,103]. In contrast, Berry et al. [104] indicated that Cu additions to FePt have no significant impact on the transition temperature, activation energy or transformation enthalpy of the chemical ordering transformation when compared to those characterizing the binary alloy. Ni additions to FePt films were also studied by Berry et al., who report an increase in the transition temperature and the activation energy for the chemical order-disorder transfor

mation with respect to Studies of the ef
mation with respect to Studies of the effects of ternary alloying additions on the FePd chemical order-disorder transformation are limited to a few investigations. Small additions (4 at %) of Cr, V and Nb to Fe were seen to diminish the degree of chemical order and the thermal stability of the L1 structure [105], as determined from a lower parameter and lower L1 volume fraction of the modified alloys in comparison with those of the unmodified one. Substitution of Pd for Pt in bulk FePd confirmed a continuous increase in the chemical order-disorder transformation temperature [106], while moderate Co or Cu additions (18 at%) to FePd thin films decreased it by as much as 50 K compared to that of the unmodified binary alloy [107–109]. The effects of Ni additions in bulk equiatomic L1FePd have been studied only by Horiuchi et al. [110] to confirm their phenomenological calculations of phase equilibria in this system. They conclude that Ni-substitution for Pd is energetically more favorable than for Fe in the L1-compound, and propose a ternary phase diagram with fcc and L139 Figure 18. Isothermal sections of Fe-Ni-Pd phase diagram at a) 650 °C b) 590 °C and c) 550 °C, as proposed by Horiuchi et al al from transformation The effect of applied external fields ( strain, magnetic field) on the chemical order-disorder transformation in ferrous systems has been studied by several authors in the last decades. To date, it

is known that external fields applied d
is known that external fields applied during or before chemical ordering can significantly alter the microstructure of the developing L1 phase, and can thus be used to tailor the final magnetic properties. It has also been established that applied external fields can enhance the kinetics of the chemiThe effect of strain on the chemical order-disorder transition of FePt, delivered through mechanical milling processing, was studied by Lyubina et al. [111]. The milling process, carried at liquid nitrogen temperatures for different hours, produced FePt alloy powders with varying crystallite sizes, diverse levels of microstrain, and different microstructures. Subsequent annealing of these powders revealed differences in the amount of L1 phase formed, suggesting that certain microstructural conditions favor the transformation. Additionally, activation energies for chemical ordering determined from these milled powders were determined to be lower than the activation energy for interdiffusion in the FePt system, and of the same order of magnitude as the activation The effect of strain on the chemical ordering of FePd alloys has been studied by either cold-working the material prior to conducting the ordering treatment [35,112,113], or by the application of stress during the chemical ordering process [114,115]. While conventionally ordered FePd alloys show coercivities of 250 mechanically processed ordered alloys exh

ibit coercivities of up to 800 Oe [35].
ibit coercivities of up to 800 Oe [35]. These enhanced values have been attributed to changes in microstructure, which considerably alter the magnetic domain stin stlly ordered polycrystalline sample, as discussed in Section 2.1.3.3, exhibits a poly-twinned microstructure, while mechanically processed samples preferentially form a monovariant microstructure without twins upon chemical ordering [115], Figure 19. The variant that is favored is the one for which the -axis is parallel to the applied stress [114]. However, a stress threshold below which a single-variant is not favored has been reported; in that case the poly-twinned microstructure with low coercivity dominates [114,115]. Figure 19. a) Microstructure of fully L1 ordered FePd alloy, obtained by annealing at 500 °C for 100 h. b) Microstructure of fully L1 ordered FePd alloy, obtained by a cold-work reduction of 97% followed by annealing at ed by annealing at . Plastic deformation delivered prior to aging of L1-forming alloys has also been observed to accelerate the chemical ordering transformation [36,43,115]. For example, et al. [115] monitored the chemical ordering transformation by electric resistance measurements in conventional and mechanically processed L1 FePd (Figure 20), and reported that the time required to fully orde from 7200 s to 720 s. The enhanced kinetics are believed to be related to the chemical ordering process being assisted by th

e simultaneous decrease in defect densit
e simultaneous decrease in defect density, in what is known as a combined reaction [36,43,113,116]. Stored strain energy incorporated in the system through mechanical processing assists heterogeneous nucleation of chemically ordered grains located at grain boundaries, dislocations, and deformation bands. The growth of these L1y mobile high-angle Figure 20. Electrical resistance measured on isothermal annealing of FePd at 843 K with or without a compressive stress of 40 MPa. A change in the electrical resistance represents the chemical disorder-order tranrder tran. Similar to the effect of strain, the application of magnetic field above a certain magnitude applied during a chemical ordering transformation has been reported to lead to preferential formation of a favorably ordered variant [115,117]. This phenomenon is clearly seen from the magnetization measurements presented by Farjami et al. [117] for single crystals of Fe-55Pd (at%) conventionally processed (Figure 21) aannealing in a magnetic field of 10 T applied in the [] direction (Figure 22). In conventional processing, the magnetization measurements of the single crystal parallel to and -directions are very similar, indicating similar amount of the three L1variants. In contrast, the sample ordered under a magnetic field applied in the -direction g magnetization results parallel to the and indicating achievement of a mono-variant structure. Favoring app

lication of a magnetic field is reported
lication of a magnetic field is reported to be especially effective at the early stages of ordering; at later stages, the growth of the L1 nuclei is dominated by other factors such as elastic interactions [117]. Figure 21. Magnetization curves for directions of a single crystal of Fe-55 at% Pd after a) ordering at 673 K for 1 h, and b) successive ordering at 43 Figure 22. Magnetization curves for x, y and z directions of a single crystal of Fe-55 at% Pd after a) ordering at 673 K for 1h under a magnetic field of 10 T, T, . 2.2 Formalism In this dissertation, thermodynamic and kinetic aspects of the A1 phase transformation in the ferromagnetic FePd and FeNi systems are discussed. This current Section presents an overview of the formalism required to understand pertinent aspects of phase transformations and ferromagnetism in material systems. Key concepts related to thermodynamic aspects of phase transitions are presented first (Section 2.2.1), including a discussion of phase equilibria and stability in the context of Gibb’s free energy. Next, the kinetics of phase transformations are di, providing general information on energy barriers involved in a phase transition anucleation and growth processes, which are relevant to chemical order-disorder phase transitions in ferrous alloy systems such as FePd and FeNi. Section 2.2.3 closes this Chapter and presents important aspects and terminology of magnetism, incl

uding a description of ferromagnetic ord
uding a description of ferromagnetic order, the hysteresis loop for ferromagnetic materials, and a brief discussion of permanent magnetic materials. A phase can be defined as a portion of a material system with homogeneous properties and composition that is physically distinguishable from other regions [27]. Phases can be considered as thermodynamic states that are characterized by a set of thermodynamic parameters, and in this context, a phase transformation is defined as a transition from one thermodynamic state to another [118]. Phase transformations occur because the initial state is unstable relative to the final state. The stability of a phase at a constant temperature () and pressure () is described TSTPG ( 2 ) Where H = Enthalpy of the system (J) The entropy ( is a measure of the randomness of the system and is the sum of the thermal entropy , which arises from the number of ways in which the available thermal energy can be shared among the conswhich arises from the number of ways in which the constituents can be distributed over a volume (V) [118]. The enthalpy (H) is a measure of the heat content in the system, and includes a contribution from the internal energy (), which relates to the kinetic and potential energy of the atoms. Kinetic energy in solid material systems arises from atomic vibration, while potential energy relates to interactions between atoms [27]. The enthalpy is then defined as

: 45 PVUTP ( 3 ) Additional energy co
: 45 PVUTP ( 3 ) Additional energy contributions may be added to the Gibbs free energy (G(P, Tto incorporate the effect of magnetic, interfacial or elastic energies. For instance, in magnetic materials subjected to a constant magnetic field , the Gibbs free energy function can be expanded to include a magnetic energy term as follows [119]: MHTSHTPG ( 4 ) Where = Applied magnetic field = Magnetization In a system of fixed mass and fixed composition at constant temperature and pressure, a phase is stable (i.e. in equilibrium) if it possesses the lowest possible value of the Gibbs free energy [27]. In mathematical terms, this definition implies that the derivative of the Gibbs free energy function is zero at stable equilibrium: ( 5 ) However, other states that do not have the lowest possible value of might also satisfy = 0; these are local minima in the Gibbs free energy that give rise to metastable equilibrium states [27]. Any state between two local Gibbs free energy minima will be unstable, as better exemplified in Figure 23. In this figure, which depicts the variation in Gibbs free energy with the system configuration, state A is the global minimum and represents the stable equilibrium state, while state B is a local minimum and represents a metastable equilibrium state. All intermediate states between A and B are unstable. 46 Figure 23. Variation of Gibbs free energy with configuration, showin

g equilibrium states at = 0. The stable
g equilibrium states at = 0. The stable state is A, while B is a metastable state. Image B is a metastable state. Image . A phase transformation ( a transformation from one equilibrium state to another), which can result as a consequence of a change in the extrinsic parameters of the system (such as temperature, pressure or magnetic field), is only possible whenever it results in a decrease in the Gibbs free energy. Therefore, a criterion for a phase transformation can be stated in terms of the difference in the Gibbs free energy between the final and initial states, which has to be negative for the transformation to be possible Where: = Free energy of final state (J) = Free energy of initial state (J) Phase transformations occurring at or near equilibrium may be classified in the context of the derivative of the Gibbs free energy with respect to a state variable, according to the Ehrenfest classification of phase transformations [120]. In this classification scheme, the transformation is said to be of th order when the derivative of the ) function with respect to temperature or pressure is the lowest derivative that shows a discontinuity at the transition temperature [66]. Accordingly, a scontinuity in the entropy or volume: ( 7 ) ( 8 ) Additionally, considering the derivative of with respect to temperature at constant pressure, it can be seen that in first-order transitions there is also a discon

tinuous change in enthalpy that is assoc
tinuous change in enthalpy that is associated with a latent heat of transformation: ( 9 ) ) are continuous in a second-order transformation, and the second derivatives show a discontinuity at the equilibrium transformation temperature. Thus, second-order phase transformations exhibit discontinuous changes in the specific heat and compressibility at the transition temperature [66]: ( 10 ) ( 11 ) Where: = Heat capacity at constant pressure (J/K) = Compressibility (mThere is no latent heat associated to a second-order phase transition, but rather an anomalous specific heat [66]. Figure 24 summarizes the thermodynamic characteristics of first and second-order phase transformations. Figure 24. Thermodynamic characteristics of a) first-order phase transformations a) first-order phase transformations . Even when there is a thermodynamic driving force ( )ormation to occur, this does not guarantee that the transformation will happen; the presence of energy barriers between metastable and stable states can limit the transformation. This concept can be better represented in Figure 25, (activation energy) needs to be overcome for the BA transformation to happen. In general, higher energy barriers lead to slower transformation rates [27]. The study of phase transformation rates belong to the realm of kinetics, which will be discussed in the 49 Figure 25. Variation of Gibbs free energy with configuration, sh

owing the driving for the transformatio
owing the driving for the transformation BA, and the activation energy that should be overcome for the transformation to occur. Image adapted from e adapted from . 2.2.2 Kinetics of Phase Transformations This Section briefly describes fundamental kinetic aspects of phase transformations, with particular emphasis on the kinetics underlying nucleation and growth processes, of relevance for the chemiormation in the FePd and FeNi systems studied in this dissertation. It was seen in Section 2.2.1 that phase transformations can be thermodynamically classified using the Ehrenfest scheme. Considering that this classifitransformations at or near equilibrium, and that many transitions in metallurgy occur away from equilibrium, other phase transformation classification schemes have been proposed, which rely on the way in which the transformation proceeds rather than on the thermodynamics involved. For instance, the classification scheme by Gibbs [30,121,122] distinguishes between continuous and discontinuous transformations. For the first case, the transformation is initiated by subtle fluctuations that are small in degree but that occur throughout large volumes, thus producing continuous changes throughout the system. Discontinuous transformation on drastic fluctuations that are small in extent, causing localized changes in the system. A similar classification was proposed by Christian [30,66,122], who differentiated

between homogeneous and heterogeneous ph
between homogeneous and heterogeneous phase transformations. In this classification, homogeneous transformations are those that occur uniformly throughout the entire system with changes occurring continuously in time as well as space. Heterogeneous transformations on the other hand, involve the spatial partitioning of the system at intermediate stages of the transformation, with macroscopically distinct regions of which some have transformed and others have not. In this sense, homogeneous transformations described by Christian are analogous to Gibbs’ continuous transformations, while heterogeneous transformations are equivalent ormations [30,123]. Transformations of the A1type, of interest for this dissertation, are classified according to the Ehrenfest scheme as thermodynamically first-order in nature [24,30,120,124]. The A1chemical ordering transformation has been reported to proceed in the FePd [10,23,33,34,41,42], FePt [23,30,41,42,111,125,126], and CoPt [31,125,127–129] systems a nucleation and growth process, for undercoolings as large as 250 K, 600 K, and 275 K for each system, respectively. At the transition temperature, the system is spatially partitioned into regions that have transformed and regions that have not, separated by an interface. Considering that this mechanism is of fundamental importance for the FePd material system studied in this dissertation, and possibly also for the FeNi system, th

e following Sections will be dedicated t
e following Sections will be dedicated to key Nucleation models of phase transformations Nucleation in solids almost always occurs at defects, interfaces or impurities, which serve as suitable nucleation sites. The free energy change associated with such a nucleation process has four main contributions, arising from: () the creation of the product phase (), () the creation of an interface (), () misfit strain energy (), ) the free energy associated with destruction of defects as they are used as ) [27]: ( 12 ) Where: = Free energy change associated to nucleation (J) = Volume of product phase nucleated (m = Interface area between product and parent phase (mFor a certain undercooling condition, there is a critical radius () of the nuclei for which the free energy change associated with the nucleation process is maximum. Below this critical radius, the system will lower its free energy by dissolution of the product the free energy will decrease if the product phase grows. The nucleation rate can be calculated from the frequency () at which each nucleus achieves a critical size multiplied by the concentration of the critical-sized nuclei. depends on how frequently a nucleus of the product phase can receive an atom from the parent phase, which depends on the surface area of the nucleus and the rate can be written as [27]: 52 ( 13 ) Where: = factor that includes vibration frequency of the atoms and the

area of the critical nucleus (1/s) = Ac
area of the critical nucleus (1/s) = Activation energy for atomic migration (J) = Boltzmann constant (J/K*mol) The concentration of critical-sized nuclei on the other hand can be described by an Arrhenius-type function which considers the nucleation barrier. Thus, the rate of nucleation takes the form: ( 14 ) Where: = Nucleation rate (nuclei/m*s) = concentration of heterogeneous nucleation sites per unit volume contained in the parent phase (atoms/m = Nucleation barrier (J) Growth models of phase transformations Growth of stable nuclei occurs by migration of atoms across the interphase interface. Phase growth can be interface-controlled, diffusion-controlled, or a combination of both (i.e. mixed control). An interface-controlled mechanism, in which atom migration across the interface is difficult, is typical of transformations in which product and parent phases are of the same composition. Diffusion-controlled growth on the other hand, typical of transformations that involve a change in composition, require long-range diffusion of atoms through the lattice [27]. Chemical order-disorder phase transformations, of interest for this dissertation, do not involve a compositional change and usually proceed through an interface-controlled mechanism [64,126]. The extent of a nucleation and growth phase transformation as a function of time during isothermal heat treatment at constant pressure is usually described

by a sigmoidal shaped curve as shown in
by a sigmoidal shaped curve as shown in Figure 26. The transformation rates are low at the beginning due to the time required for stable nuclei to form and begin growing ( in Figure 26), high at intermediate times, and low again afterwards as the un-transformed volume becomes less available. The fraction transformed can be described by the Johnson-Mehl-Avrami-Kolmogorov (JMAK) formulation, which considers different nucleation conditions and growth mechanisms [30,66]. In its most general form, the JMAK model is as follows [27,130]: ( 15 ) Where: = Fraction transformed varying from 0 to 1 (dimensionless) rate constant (1/s) = time (s) = Avrami exponent, which is typically independent of temperature (dimensionless) 54 ( 16 ) The Avrami exponent depends on the nucleation and growth conditions, with the physical values for each condition presented in Table 1: ( 17 ) Table 1. Numerical values corresponding to nucleation and growth conditions that determine the Avrami exponent. Adapted from . Nucleation condition Growth dimensionality Growth mechanism 0: Site-saturated nucleation * 1: 1-D growth ½: diffusion-controlled 1: constant nucleation rate 2: 2-D growth 1: interface-controlled 3: 3-D growth * Site saturated nucleation refers to a condgrowth begins Figure 26. Characteristic sigmoidal curve associated with isothermal nucleation and growth processes. Image adapted from Key Concepts in Magnet

ism In this dissertation, FePd and FeNi
ism In this dissertation, FePd and FeNi alloys that undergo a chemical order-disorder transformation are studied. These alloys are ferromagnetic at room temperature, and thus an important part of this dissertation deals with the characterization of intrinsic and extrinsic magnetic properties, as well as with determination of the critical temperature at which the alloys become paramagnetic, known as the Curie temperature. This Section presents basic aspects and key concepts in magnetism that are necessary for understanding the magnetic data prThe magnetic moment in an atom is a vector quantity associated with the spin of each electron and its orbital motion around the nucleus [132]. As atoms come together to build a crystal, interactions among them may lead to an orderly arrangement of the magnetic moments; in the case of ferromagnets, a parallel alignment of the moments occurs. The interaction between neighboring atoms is a purely quantum mechanical ( 18 ) Where: = Exchange interaction iSThe exchange interaction is generalized to sum over all pairs of atoms within a lattice. The exchange constant depends on, among other factors, the distance between the atoms being considered [134]. � 0 indicates a ferromagnetic interaction, which tends to align spins in a parallel manner yielding a spontaneous magnetization in the absence of an applied field. As the temperature increases, thermal agitation leads

to gradual disorder of this parallel al
to gradual disorder of this parallel alignment and to a corresponding reduction in the magnetization [135]. At a critical temperature, known as Curie temperature , the individual magnetic moments no longer interact strongly and the spontaneous magnetization of the crystal vanishes in the absence of an applied field, that is, the material becomes paramagnetic. Magnetic phase transitions, such as the ferromagneticparamagnetic transformation described above, can be described by the same general principles presented in Section 2.2.1 for phase transformations. In this case, the Gibbs free energy contains an additional term () (as was seen in Section 2.2.1), and a thermodynamic classification into first and second-order makes use of the derivatives of the Gibbs free energy. At a constant pressure, a first-order magnetic transition exhibits a discontinuity in the entropy and the magnetization [119]: ( 19 ) ( 20 ) In a second-order magnetic transformation, the first derivatives of the Gibbs free energy are continuous, but there is a discontinuity in the second derivatives, that is, in the specific heat and isothermal susceptibility [119]: ( 21 ) 57 ( 22 ) The ferromagneticparamagnetic transition is usually a second-order transition, with gradual changes in magnetization at the Curie temperature, but a discontinuous ferromagnetic and paramagnetic states [119]. The magnetic response of a ferromagnetic mate

rial to an applied magnetic field can be
rial to an applied magnetic field can be explained in the context of magnetic domains, small regions in the material which are spontaneously and uniformly magnetized. These regions are separated by domain walls. In the demagnetized state (zero field), the direction of magnetization varies from one domain to another in such a way that the material has zero net magnetization [133]. At small applied fields, domains aligned favorably with respect to the field grow at the expense of others through a process of domain wall motion, increasing the overall net magnetization of the sample. At moderate applied fields, domain rotation processes tend to operate, in which the atomic magnetic moments in unfavorably aligned domains rotate from their original direction into a crystallographic easy-axis nearest to the direction of the applied field. At yet larger fields there is a gradual coherent rotation of the magnetic moments into the direction of the applied field [136]. In the final state, the sample is in a single-domain state, and the highest possible value of magnetization is achieved, referred to as magnetic saturation (). Removal of the magnetic field will not cause the magnetic polarization to relax back to its original state but the material will retain some remnant magnetization, . Demagnetization of the material is achieved when applying a field in the opposite direction with a magnitude known as the coercive field

, . If the material is driven to magnet
, . If the material is driven to magnetic saturation in the negative direction, and then again in the positive direction, a hysteresis loop is traced as shown in Figure 27a. Magnetic properties as are intrinsic, meaning they are material-specific and are independent of the sample’s mass or microstructure. These quantities are determined by the atomic origins of magnetism, and they are realized at few interatomic distances [137]. Properties as remanence () and coercivity () are extrinsic, and they vary with the size, shape and microstructure of the sample [8]. Figure 27. a) Typical ) curve of a ferromagnet, showing saturation magnetization , and coercivity . Magnetic domains at different stages of the hysteresis loop are included for illustration purposes. Image adapted from purposes. Image adapted from . b) Hysteresis loops for soft and hard ferromagnets. Image adapted from The shape of the hysteresis loop is strongly affected by the presence of preferential directions for the alignment of the magnetic moments in the absence of a magnetic field, that is, if there is magnetic anisotropy [133]magnetic anisotropy, but of particular interest to this dissertation is the magnetocrystalline anisotropy, which descrispontaneous magnetization to lie along particular crystallographic directions, known as easy axes [133,136]. In this case, an anisotropy energy () is required to direct the magnetization vectors away f

rom the easy axis direction. The anisotr
rom the easy axis direction. The anisotropy energy is approximated in a uniaxial system as a power series, but often only the first term of the the 21sinKEA ( 23 ) Where: d magnetization vector (degree) The origin of magnetocrystalline anisotropy is the quantum-mechanical spin-orbit interaction [133]. In ferromagnetic alloys, chemical order also contributes greatly to the magnetocrystalline anisotropy, as has been demonstrated by theoretical [138–142] and experimental [143–145] studies carried out on L1 chemically ordered ferromagnetic systems, of interest to this dissertation. In AB alloys with the L1 crystal structure, in which the constituent A and B atoms are stacked in alternating layers along a tetragonal -axis, there is an easy axis of magnetization parallel to the stacking direction. The system thus exhibits a uniaxial magnetocrystalline anisotropy, with an anisotropy energy which is directly related to the degree of long-range chemical order Experimental investigations by Kamp et al. [144] on L1 FePd reveal a rather linear increase in with variations in . Theoretical analyses on L1 FePt by Ostanin et al.[140] yield a quadratic dependence of on the amplitude of the chemical concentration waves, which represents the chemical order parameter. Other first-principle theoretical investigations have indicated that additional factors such

as the tetragonality of the L1crystal s
as the tetragonality of the L1crystal structure [138–141] and the composition [142] have a minor but definite Figure 28a presents theoretical results obtained by Kota et al. [141], where the variation in for ferromagnetic FePt with the L1-type structure is shown as a function of both increased and ratio. It is seen that is primarily dependent on chemical order, and that there is a small linear dependence of on the magnitude of the ratio. Figure 28b shows theoretical results obtained by Aas et al. [142] of the for ferromagnetic L1 FePt of varying composition as a function of increased . An increase in the magnetocrystalline anisotropy energy is observed with increased Fe content in the presence of some degree of chemical disorder. These results, corroborated by experimental findings reported by Barmak et al. [145], suggest that for varying composition and constant degree of chemical order, the degree to which Fe atoms fill the nominal Fe-layers dominates Figure 28. First-principle calculation results of the variation in the magnetocrystalline anisotropy energy of L1 type FePt as a function of a) c/a ratio and long-range order parameter tio and long-range order parameter , and b) long-range , and b) long-range . Ferromagnetic materials with uniaxial magnetocrystalline anisotropy, as L1 FePd FeNi, have the potential to exhibit a high resistance to demagnetization that results in a hard-magnetic behavior

. Hard ferromagnets require a high appli
. Hard ferromagnets require a high applied field to achieve saturation and their hysteresis loop is very broad. The opposite case is a soft ferromagnet, in which saturation is achieved teresis behavior is observed [6,133,134] (Figure 27b). Hard ferromagnets are commonly used as permanent magnets for energy generation and transformation. A good permanent magnet is one that has a high value of saturation magnetization , a high value of remanence coercivity that are stable at the operation temperature. Under these circumstances, the material is able to store a large amount of magnetic energy, which is commonly quantified by a figure of merit referred to as the energy density (The energy density isdefined as the maximum product of the flux density magnetic field manifest in the second quadrant of the hysteresis loop [134]. Permanent magnets functionalize many types of advanced technologies that support alternative energy systems, communications and military applications, and consumer devices. Current advanced permanent magnets, with energy densities in the range 18-52 MGOe, contain rare-earth elements as neodymium, samarium and dysprosium (Figure 29). Recent rare-earth availability constraints, imposed by geopolitical factors, led the international scientific and technological communities to recognize the need to realize novel hard-magnetic materials made of earth-abundant elements to support the global energy

challenge [5,7,8]. Among the candidates
challenge [5,7,8]. Among the candidates for the development of next-generation permanent magnets, high-magnetocrystalline anisotropy -structured ferrous materials with a high saturation magnetization exhibit theoretical energy densities superior to SmCo magnets and comparable to NdB magnets, Table 2. The costliness of Pd/Pt limits the use of L1 FePd and FePt for bulk applications, but the low cost and good availability of Fe/Ni make L1 FeNi a very promising material for permanent magnets. Figure 29. Permanent magnet performance, as quantified by the energy density max in MGOe, versus average selling price per kg. Sintered rare earth based magnets exhibit high energy densities, up to 52 MGOe. Image MGOe. Image . Table 2. Comparison of magnetic properties for L1 ferrous compounds and common rare-earth magnetic materials terials . K (erg/cm3) MS (emu/cm3) TC (K) max (MGOe) (Theoretical) FeNi 1.3x10 1270 830 42 FePd 1.8x10 1100 760 48 FePt 6.6 x10 1140 750 51 B 4.6 x10 1270 585 64 SmCo 11-20 x10 910 1000 33 EXPERIMENTAL TECHNIQUES The overarching goal of this research is to establish the effects of intrinsic and extrinsic modification on the A1 phase transformation in the model FePd system, with the goal to clarify kinetic and thermodynamic pathways to achieve formation of tetrataenite (L1 FeNi). To that end, three AimsUnderstand the effects of intrinsic modification, ternary alloying addition, on the ph

ase transformation character, structure
ase transformation character, structure and magnetic properties of L1Understand the effects of extrinsic modification, plastic deformation, on the phase transformation in FePd. Evaluate the effects of ternary alloying additions and plastic deformation on the transformation in FeNi Studies related to Aim 1 employed bulk FePd alloys modified with ternary additions in the general formula Fe )of Ni substituting for Pd were selected to approach the L1 FeNi phase, while simultaneously allowing studies of the effects of Ni additions on the structure, magnetism and phase formation character of FePd. Additions of Cu were selected on the basis of the similarity between Ni and Cu in terms of atomic size, electronegativity, and structure into which they crystallize, while the magnetic order characteristics at room temperature are different for both elements. Thus, additions of these elements are anticipated to provide a better understanding of the role of magnetism in the stability of the L1 phase. The ternary alloys were synthesized by arc-melting and were subsequently annealed below the chemical order-disorder temperature to promote the formation of the L1 phase. Structural characterization through X-ray diffraction and magnetic characterization through Vibrating Sample Magnetometry was pursued before anning Calorimetry was used to study phase transformations in the annealed samples. Comparison of the results obtaine

d for ternary alloys to those obtained f
d for ternary alloys to those obtained for unmodified binary FePd allows relations to be made between ternary alloying additions and properties of the A1 and L1 phases, as well as on the transformation. Studies pertaining Aim 2 of this dissertation used bulk FePd synthesized by drop-casting and annealed below the chemical order-disorder temperature to promote the formation of the L1 phase. Part of the annealed alloy was cold-rolled in several steps to deliver a large amount of strain. X-ray diffraction and Differential Scanning Calorimetry techniques were used to characterize the alloy before and after cold-rolling, to quantify the relationship between plastic deformation energy and chemical disorder. Moreover, considering that plastic deformation processes introduce a high density of crystalline defects in the material which may have an imwere undertaken to evaluate the effect of cold-rolling on the A1 chemical order transformation. Isothermal annealing in situ in the DSC at different temperatures to monitor the A1 transformation in undeformed and in cold-rolled samples provided insight on the different nucleation and growth mechanisms and the energy barriers for the transformation. Structural characterization through X-ray diffraction performed on isothermally annealed samples allowed determination of the effect of plastic deformation and annealing temperature on the structural characteristics of the resulta

nt L1Finally, Aim 3 evaluates ternary a
nt L1Finally, Aim 3 evaluates ternary alloying and severe plastic deformation as thermodynamic and kinetic routes to access the elusive L1 FeNi phase. Ternary additions of Al, V, and Ti (2 at%) were selected on the basis of their negative heats of formation for an L1 phase with Fe or Ni [147]. Selected samples were fabricated with Ni instead of natural Ni, to enhance the neutron contrast with Fe and allow the investigation of L1 chemical order in these samples. Severe plastic deformation by cold-rolling and cryomilling was used to deliver large amounts of strain and increased number of extrinsic structural defects anticipated to favor atomic diffusion. Subsequent annealing protocols below the chemical order-disorder temperature were used to promote L1chemical ordering. Advanced structural characterization techniques such as synchrotron lution time-of-flight neutron diffraction, together with magnetic characterization by magnetometry, allow relations to be made between ternary alloying and severe plastic deformation processing on the structural and magnetic This Chapter provides a description of the experimental techniques used in this dissertation for the synthesis, processing, and characterization of all samples. Section 3.1 provides a summary list of all samples studied, and describes the arc-melting and drop-e processing techniques used throughout this dissertation to accomplish structural, microstructural a

nd magnetic changes. These include melt-
nd magnetic changes. These include melt-spinning, cold-rolling, cryomilling, and annealing. Section 3.3 provides a description of the instruments used in this dissertation for the characterization of the samples, including scattering techniques, magnetometry, and calorimetry methods. The operating principles are provided and data analysis procedures are presented. Specific conditions used for sample synthesis, processing and characterization are not included here, as they vary from sample to sample. Instead, details are provided for samples of each specific Aim in Chapter 4. This Section describes the experimental techniques used for the preparation of the FePd and FeNi-based alloys studied in this dissertation; a complete list of all samples examined is provided in Table 3. In summary, FePd samples with ternary alloying additions of Cu and Ni were used to evaluate the effect of chemical modification on L1phase formation. Additionally, binary FePd alloys processed through different techniques were employed to gain a better understanding of the effect of plastic deformation on the phase transformation. Finally, FeNi binary alloys and FeNi with small substitutional additions of Ti, V and Al were processed by different techniques in pursuit of the L1 structure. The samples have been grouped according to the specific Aim to Table 3. List of samples studied in this dissertation. Synthesis methods Processing method

s Aim 1: Understand the effect of terna
s Aim 1: Understand the effect of ternary addition on L1phase formation Isothermal annealing in 50- = 3, 5, 7 at%) 50- = 3, 5, 7 at%) Aim 2: Understand the effects of plastic phase Drop-casting Isothermal annealing transformation Isothermal annealing Aim 3: Evaluate the potential of ternary alloying and plastic phase formation in FeNi drop-casting Cryomilling or cold-Isothermal annealing 0.50.5Isothermal annealing 0.50.5Most samples used in this dissertation were synthesized in bulk form by an arc-melting process starting from high purity elements. During arc-melting, a DC-current is applied between a tungsten tip and a water-cooled copper plate under an argon atmosphere, producing an electric arc that generates a plasma that reaches temperatures up to 3500 °C.The material of interest, sitting on top of the copper plate, is melted by the plasma. Once the DC current is removed and the melt is allowed to cool down, an ingot of the desired composition is obtained. To prevent contamination by oxidation, an oxygen getter is placed in the chamber and is melted prior to the material of interest. Compositional homogeneity can be achieved through repetition of the arc-melting process. In this dissertation, an Edmund Buhler MAM-1 mini-arc melting system operated in an argon atmosphere was used, and at least three consecutive repetitions of the arc-melting process were performed to ensure homogeneity of the FePd-

based and FeNi-based samples. Selecte
based and FeNi-based samples. Selected FePd and FeNi-based alloys used in this dissertation were fabricated by drop-casting. In this technique, the elements are melted to produce an alloy which is then allowed to fall by gravity into a mold of a desired shape. This technique was carried out at the Materials Processing Center (MPC) at the U.S. DOE Ames Laboratory in Ames, Iowa. It was used to fabricate ~50 g cylindrical samples of FePd, FeNi and FeNiTi alloys of approximate dimensions 10 cm. Ingots produced by any of the different techniques described in Section 3.1 were subsequently processed by a variety of methods that are described in this Section. These techniques were used either to facilitate further processing of the samples, to impart severe plastic deformation to the samples, or to promote the formation of a to facilitate further processing Melt-spinning is a rapid solidification technique (cooling rate ~ 10 K/s) which is commonly used to achieve non-equilibrium conditions under which interesting microstructures, metastable phases and enhanced mechanical or structural properties can be obtained [148]. In this dissertation however, melt-spinning was used for none of the reasons described above, but to render FeNi-based alloys amenable to cryomilling. In the melt-spinning process, an ingot of material is placed inside a crucible containing a small orifice at the bottom. The crucible goes inside induct

ion coils, below which there is a rotati
ion coils, below which there is a rotating water-cooled copper wheel. An inert atmosphere is guaranteed in the melt-spinning chamber by evacuation and backfilling with argon. RF-induction is used to heat and melt the material, after which an argon backpressure is used to eject the melt through the crucible’s orifice onto with the wheel, the melt solidifies and produces ribbons which are thrown off to a collection sleeve. The orifice size, orifice-wheel distance, wheel speed and applied pressure can be varied to achieve different flow dynamics of the molten stream, thus affecting the final product. In this dissertation, an Edmund Bühler GmbH) melt-spinner operated in an argon atmosphere was used to obtain FeNi-based alloys that were subsequently subjected to cryomilling. Quartz or boron nitride crucibles set at a distance of 3-8 mm to the Cu wheel were used. The effective rotating tangential speed of the wheel was set at 31 m/s. Processing of FeNi-based alloys through cryomilling Mechanical milling is a high-energy processing technique in which materials, generally in powder form, are broken into smaller particles. The process takes place inside a vial that contains both the material to be ground and grinding elements. The grinding elements are forced to move inside the vial, impacting the material and providing the milling action. During the impact, the material experiences compressive, shear or impact stresse

s that result in heavy plastic deformati
s that result in heavy plastic deformation, as manifested by the presence of a variety of crystal defects such as dislocations, vacancies, stacking faults and increased number of grain boundaries [149]. The milling process is accompanied by an increase in overall temperature, caused by the multiple collisions taking place. If the temperature rise is significant, this can promote phase transformations or lead to dynamic ocesses in metals [149–151]. To avoid such processes, mechanical milling can be performed at cryogenic temperatures, in which case it is referred to as “cryomilling”. In this dissertation, cryomilling was used to process selected FeNi-based alloys. The equipment used, a SPEX SamplePrep 6770 Freezer/Mill, is composed of a magnetically driven stainless steel impactor that moves back and forth inside a stainless steel vial that contains the material to be ground. During the milling process, the vial is submerged in liquid nitrogen to minimize the temperature rise. Prior to milling, the vial is loaded with the material and sealed inside a glovebox to guarantee an inert atmosphere during the process. Cryomilling was used in this dissertation to deliver a high density of crystalline defects to FeNi-based alloys. Mechanical milling processes are known to produce large concentrations of lattice defects [149], including vacancies, in amounts comparable to those found near the materials’ melting temperatu

res [152]. These defects can deliver a p
res [152]. These defects can deliver a positive effect on atomic diffusion [27,153]. Thus, cryomilling is envisioned as a potential technique to favor the kinetically limited L1 FeNi phase formation. Furthermore, by performing the process at cryogenic temperatures, the occurrence of thermally-driven recovery and recrystallization processes in the highly deformed Processing of FePd and FeNi-based alloys through cold-rolling Cold-rolling is a common technique for achieving severe plastic deformation of materials. During cold-rolling, the material is forced to pass through the gap between a pair of rollers. The size of the gap determines the final thickness achieved. Usually, small consecutive step reductions are used to achieve a high percentage of cold work (%CWdefined in terms of the initial () and final (71 100%00 ( 24 ) As a result of the cold-working procedure, crystalline defects such as a high density of dislocations are achieved, accompanied by the development of a crystallographic texture in the direction of rolling [154,155]. In this dissertation, a Stanat Rolling Mill (Ames Lab) was used for cold-rolling of FePd and FeNi-based alloys, to deliver large amount of strain and defects that could affect the A1 transformation in the systems studied. The process was performed ng the same rolling direction. The applied load varied in the oys to promote phase transformations Section 2.1.2.1 described phase eq

uilibria of the Fe-Pd system, subject of
uilibria of the Fe-Pd system, subject of the present study. It was seen that a chemically ordered L1 structure is stable at low temperatures for the near-equiatomic compositions. Yet, when cooling a molten FePd alloy to room temperature, typically the high-temperature A1 phase is metastably retained. This happens whenever the cooling rate is not sufficiently slow to allow the transformation to proceed. In this case, an annealing protocol below the order-disorder transition temperature is required to obtain the equilibrium L1 phase in FePd. In the FeNi system, it is known that an annealing step below is not sufficient to achieve phase in laboratory time scales. Nonetheless, it is anticipated that to promote L1chemical order in this system, an annealing step will be necessary in conjunction with hance diffusion at low temperatures. In this dissertation, annealing processes were employed to either: homogenization of a particular alloy at high temperatures, or ) promote L1 phase formation. Annealing was performed in a tube furnace, and details of annealing temperatures and times, specific to each sample, are provided in the Results Section. To prevent oxidation of the samples, these were encapsulated in evacuated (1 x 10Tantalum foil used asChemical modification through substitutional ternary alloying additions and microstructural modification through severe plastic deformation result in changes in the structura

l characteristics and magnetic propertie
l characteristics and magnetic properties of the FePd and FeNi-based alloys studied in this dissertation. These modifications also have a significant effect on the thermodynamics and kinetics of the A1 chemical order-disorder transformation. In this Section, the different characterization techniques used to monitor changes in the structure, magnetism, and phase transformation character are described. Procedures for the analysis of the data obtained from these experimental techniques is also provided. Elemental analysis and compositional homogeneity determination of the as-arc-melted, as-drop-cast, and as-melt-spun samples of this dissertation was performed either through energy dispersive X-ray spectroscopy coupled to a Hitachi S4800 field emission scanning electron microscope (SEM-EDS) or through X-ray fluorescence (XRF, Bruker M4 Tornado, Ames Lab). These techniques are based on the principle that each element has a unique electronic structure that allows emission of characteristic X-rays. The sample is bombarded by a high-energy beam of electrons (EDS) or by a beam of short-wavelength X-rays (XRF), exciting core-level electrons to empty states and leaving the atom in an ionized state [156]. The system then relaxes by filling the core hole with an electron of lower binding energy, producing an X-ray photon of energy equal to the difference between the two shells involved. Since the energy levels of the vario

us shells are related to the number of c
us shells are related to the number of charges in the nucleus, the energy of the X-ray produced is characteristic of the element, and can thus be used to obtain a local chemical ical The accuracy that can be obtained with EDS and XRF techniques depends on whether standards are used as reference prior to measurements (standardized analysis) or not (standardless analysis) [157]. Standardless analyses make use of atomic number, absorption and fluorescence (ZAF) correction factors built into the software, and are common when the nominal composition of the material is known and a simple verification is required. In this dissertation, alloys of a desired composition were synthesized from elemental materials that were carefully weighted before synthesis. A standardless analysis performed on at least three different locations of the samples was used to determine the average composition. Considering that all alloys characterized through EDS or XRF were single-phase, and that the measured compositions in all cases were very close to the nominal values, these measurements were considered reliable despite the fact that no standards were used. Structural characterization of the alloys studied in this dissertation was accomplished through laboratory and synchrotron-based X-ray diffraction and neutron diffraction. Diffraction experiments were employed for phase analysis and investigation of lattice parameters, microstrain (),

and size of coherently diffracting doma
and size of coherently diffracting domains (crystallite size, ). Additionally, for chemically ordered samples, diffraction experiments were used to quantify the degree of long-range chemical order. In a diffraction experiment, the structure of a material is probed with mono-chromatic radiation with a wavelength similar to the interatomic distances of interest, in this particular case, X-rays or neutrons. An X-ray/neutron beam incident on a crystalline sample is scattered by the atoms in the sample in different directions. Although the scattered X-rays cancel in most directions through destructive interference, in some instances they interfere constructively producing a scattered beam. Bragg’s law relates the angle at which the scattered beam is detected with the spacing between diffracting sin2dn ( 25 ) Where: = Interplanar spacing (Å) = Angle of incidence (or scattering) of X-rays (degrees) = Integer number (dimensionless) ) of the beam and colleca diffraction pattern is obtained comprised of Bragg peaks that correspond to different crystallographic planes. A common diffraction experiment is X-ray powder diffraction performed in a Bragg-Brentano () configuration (Figure 30). In this setup, used extensively in this dissertation, the material sboth the X-ray tube and detector are rotated around it at angles of with respect to the sample’s surface. The diffraction pattern obtained is plotted as a functio

n of (angle between incident and diffrac
n of (angle between incident and diffracted beam), and a least square analysis of the positions of the Bragg peaks obtained yields the lattice parameters of by a crystal. For X-rays tothe distance ABC needs to be an integer number of the wavelength of the X-rays. From geometric construction, the distance ABC is equal to 2Additional to Bragg peak positions, the breadth of the diffraction peaks also provides useful information on the structural characteristics of the sample, such as the size of coherently diffracting domains ( crystallite size) and microstrain. For nanostructured systems, only a small amount of planes contribute to X-ray diffraction. In this case, total destructive interference of X-rays that do not entirely satisfy the Bragg condition is not achieved, and this results in broadening of Bragg peaks [158]. Another contribution to peak broadening is the presence of anisotropic strain, commonly referred to as microstrain. In this event, Bragg’s condi76 Figure 31. Comparison of X-ray diffraction under ideal conditions and with [160]Numerous methods for line-broadening analysis are available, the most common of which is the Williamson-Hall approach [161]. In this method, peak breadth (after instrumental broadening correction, see Appendix 1) is assumed to be a result of the sum of crystallite size and microstrain contributions. The full-width at half-maximum () of

( 26 ) Where: (rad/Å) / = microstrain (dimensionless) The intercept and slope of a linear fit to the data points provide an estimation of the crystallite size and microstrain, respectively. To account for elastic anisotropy effects, modified versions of the Williamson- Hall approach can be used, which consider elastic elastic ( 27 ) Where: = uniform stress. Relates to = = Young’s moduli perpendicular to planes (Pa) In this dissertation, line broadening analysis was performed through the modified Williamson-Hall technique. The Bragg peaks in the diffractograms were fit with a pseudo-Voigt function to extract the full-width at half-maximum. Correction for instrumental contributions was performed according to Appendix 1. The elastic moduli considered for the analysis were calculated from the elastic constants reported for fcc- from the elastic constants reported for fcc-50Pd50 and L10-type Fe [164]. The integrated intensity of the Bragg reflections in a diffractogram also provides useful information for structural analysis. In case of multi-phase coexistence, a relation of the integrated intensity of peaks corresponding to different phases can be used to obtain the phase volume fraction. For instance, in a two-phase + system, the following relation provides an estimate of the relative volum

e fractions [165]:
e fractions [165]: ( 28 ) Where: = Experimental integrated intensity of the main Bragg peak of the phases, respectively (arbitrary units) = Theoretical integrated intensity of the main Bragg peak of the phases, respectively (arbitrary units) (see In a system that undergoes the A1 transition nucleation and growth, the degree of transformation will depend on both temperature and time [31]. The resultant sample may be inhomogeneous (consisting of both L1 and A1 regions) if sufficient time is not allowed for the transformation to go to completion. In this case, the diffraction pattern will contain Bragg reflections for both L1 and A1 phases [47]. This is particularly evident for reflections that undergo peak-splitting upon chemical ordering, for which three peaks can be observed (one for the cubic A1 phase, in-between the two split reflections of the L1 phase). A straightforward calculation of the L1 fraction (uses a relation of the integrated intensities of Bragg reflections involved in peak-splitting. The procedure is described by Cebollada et al. [47] and is exemplified in the following 200) fundamental Br00001)002(1)200(1)200(1)002(1)200(0LLALLIIIIIf ( 29 ) Where: (200)L1 = Integrated intensity of the ( reflection (arbitrary units) (002)L1 = Integrated intensity of the ((200)A1 = Integrated intensity of the (Furthermore, in L1-st

ructured systems the degree of chemical
ructured systems the degree of chemical order as long-range order parameter can be determined from a relation of the integrated intensity of a superlattice peak to the integrated intensity of a fundamental peak. The procedure involves comparison of the experimentally obtained intensities with theoretical values predicted for a fully chemically-ordered sample [81]: Where: = Integrated intensity of a superstructure and fundamental line, respectively(arbitrary units) = Theoretical intensity of a superstructure and fundamental line, respectively (arbitrary units) (see Appeffraction differ, mainly, in the process of X-ray generation and in the intensity of the resultant X-ray beam. For laboratory X-ray diffraction, a filament in an X-ray tube is heated to produce electrons, which are accelerated towards a metal target (typically Cu, Mo, Co or Fe) when applying a voltage. Upon collision, the electrons dislodge inner-shell electrons from the target, and in this process X-rays are produced [81]. In contrast, at a synchrotron faciwith a circumference of about one kilometer contains electrons that travel close to the speed of light. The electrons are bent around the ring with bending magnets or magnetic insertion devices, and in this process they lose energy by generating synchrotron radiation [166]. This process allows for energy tunability of the X-rays produced, as well as for much higher fluxes than those obtained

in a laboratory equipment. In this dis
in a laboratory equipment. In this dissertation, both laboratory and synchrotron-based X-ray diffraction experiments were performed for structural analysis of the FePd and FeNi-based materials. The laboratory diffractometers used included PANalytical and Rigaku instruments with a 80 copper target ( = 1.54056 Å), and a PANalytical instrument with a cobalt target = 1.789010 Å) (Ames Laboratory). Synchrotron X-ray experiments were performed employing a two-circle diffractometer with a Si() monochromator at the Beamline X16C at the National Synchrotron Light Source (NSLS) (Brookhaven National In addition to laboratory and synchrotron-based X-ray diffraction, neutron diffraction experiments were used in this dissertation as an additional technique for structural characterization. The neutron diffraction technique is similar to X-ray diffraction, but gives complementary information; while X-rays interact primarily with the electron cloud surrounding each atom, neutrons interact directly with the nucleus [167]. As a result, the diffracted intensity for X-rays is larger for atoms with large atomic number, while the diffracted intensity for neutrons is different for each isotope, with no direct correlation with the atomic number er e very similar X-ray cattering lengths of natural Fe and Ni [81,89], which make the detection of chemical ordering in the FeNi system extremely difficult as demonstrated in Section 2.1.3

.5, neutron isotopic contrast is of grea
.5, neutron isotopic contrast is of great use in the study of FeNi alloys. In this dissertation, isotopic Ni was used for the synthesis of selected samples to enhance any nuclear scattering signal originating from long-range chemical ordering. Neutron powder diffraction experiments were performed at the High Resolution Powder Diffraction (HRPD) beamline at the ISIS pulsed neutron source facility (Rutherford Appleton Laboratory, UK). At the ISIS pulsed neutron source, diffractometers operate in a time-of-flight mode, in which a pulsed polychromatic neutron beam travels through a known flight path at the end of which it strikes the sample. The time-of-flight to the sample is registered, from which the energy of the scattered neutrons (detected in an array of detectors surrounding the sample) is determined. From the neutron energies and -spacing information from the sample is obtained according to Bragg’s law. The HRPD beamline is the highest resolution neutron diffractometer of its type in the world, providing the means to detect very subtle structural details, such as the small tetragonal distortion characteristic of L1Magnetic characterization: magnetometry Characterization of intrinsic and extrinsic magnetic properties of the synthesized and processed materials studied in this dissertation were carried out in a vibrating sample magnetometer (VSM, Quantum Design Versalab) or in a Superconducting Quantum Interfe

rence Device magnetometer (SQUID, Quantu
rence Device magnetometer (SQUID, Quantum Design MPMS). Both instruments provide a volume-averaged magnetization measurement, but differ on their sensitivity and the method used to generate the magnetic field and to detect the samples’ magnetic moment. In general terms, the SQUID magnetometer has better sensitivity than the VSM magnetometer, but the latter is commonly used for measurements of bulk ferromagnets which exhibit large magnetic moments [132]. The VSM equipment (Figure 32) consists of three main systems: (generation of a magnetic field, () pick-up coils for detection of the magnetic moment of the sample, and (ng the entire measurement, the sample is vibrated by a linear motor at a frequency an amplitude selected by the operator. A solenoid, usually made of a superconducting material, is used to produce a uniform magnetic field along its axial cylindrical axis [133]. For the particular case of the Quantum Design Versalab VSM, used in this dissertation, a NbTi solenoid allows generation of magnetic fields up to 30 kOe. As a result of the applied field, a magnetic moment may be induced in the sample being measured. The oscillating magnetic flux of the sample induces a voltage in the pick-up coils, which is proportional to the magnetic moment of the sample. The working temperature of the VSM must be established below the critical temperature at which the solenoid material becomes superconducting. To this

end, a cooling system is used, usually
end, a cooling system is used, usually involving cryogens as liquid helium. In the case of the Quantum Desing Verslab VSM, the cooling system is cryogen-free, and consists of a compressor and a cold head connected in a closed system. The compressor provides helium gas at a high pressure which expands in the cold head, accomplishing the required cooling. The cooling system can also be used for measurements below room temperature, as low as 50 K. The oscillating motion is achieved by using a linear motor. Image adapted The SQUID magnetometer is also composed of a superconducting magnet and a detection coil, but in addition, contains a superconducting quantum interference device which is connected to the detection coils and serves as an extremely sensitive current to voltage converter [136]. Current variations in the detection coils produce a corresponding variation in the SQUID output voltage, proportional to the magnetic moment of the sample. For the Quantum Design MPMS-XL5 SQUID system used in this dissertation, magnetic fields up to 50 kOe can be produced, and temperature can be varied between 2 In this dissertation, VSM and SQUID instruments were used to obtain measurements of the magnetization as a function of applied field , either at room temperature or at 10 K. The internal field experience by the sample was determined (arising from sample geometry) from the The demagnetizing field was calculated as

is the magnetization and is the demagn
is the magnetization and is the demagnetizing factor, which can be obtained from tabulated values for the geometry under consideration [133] or by measuring a standard with the same geometry as the sample of interest. Form the vs. data obtained, the saturation magnetization was extracted directly in saturated samples, and an estimate was obtained through extrapolation by the Law of Approach to Saturation in unsaturated samples. This law models the high field portion of the initial magnetization curve as [168]: ( 31 ) Where: = saturation magnetization = term related to the existence of structural defects/magnetic inclusions = term related to magnetocrystalline anisotropy = high field susceptibility As shown by Grössinger [169], graphic approaches can be used to obtain the constants of the Law of Approach to Saturation. For instance, by plotting versus 1/the intercept of the linear part can provide an estimate of . This procedure was used in this dissertation to get a rough estimate of for samples that did not achieve saturation Phase transitions in closed systems at a fixed pressure are closely related to changes in enthalpy and/or heat capacity of the material [27,170]. Consequently, thermo-analytical techniques such as differential scanning calorimetry (Duseful in the study of phase transformations. In this dissertation, different structural and magnetic phase transformations in FePd and FeNi-ba

sed systems in the temperature range 300
sed systems in the temperature range 300 K 700 K were studied using a simultaneous thermal analyzer (STA 449 F3 Jupiter, Netzsch), which can perform coupled DSC and thermogravimetry (TG) measurements, evaluating both caloric effects and mass changes as a function of temperature and/or time (Figure 33). 85 Figure 33. Schematic representation of Simultaneous Thermal Analysis (STA). TA). . The DSC component of the STA consists of an inert reference and the sample under consideration, placed inside individual pans that sit next to each other in a furnace. Enthalpy or heat capacity changes in the sample lead to a temperature difference relative to the inert reference, which is registered by thermocouples and translated to heat flow information by the use of calibration experiments. The TG component of the STA consists of a balance that is coupled to the sample carrier and that registers mass changes simultaneous with the DSC operation. In this dissertation, the DSC was used extensively to study phase transformations, while TG analysis served to monitor possible mass changes associated with sample oxidation. Calorimetry experiments under isothermal conditions were used to study the kinetics of the chemical ordering transformation in FePd samples, and isochronal experiments ( constant heating rate, 20 K/min) were used to study the order-disorder phase transformation and magnetic transitions in FePd-based systems.

Fist-order phase transformations, which
Fist-order phase transformations, which require heat (endothermic) or evolve heat (exothermic), appear in DSC as sharp peaks. The onset temperature of the peak defines the beginning of the transition and the area under the peak represents the enthalpy of transformation. Phase transformations for which there is no latent heat but rather a change in heat capacity are observed as a shift in the DSC signal rather than a sharp peak [172], and can be better identified as a discontinuity in the derivative of the DSC signal as a function of temperature. The Netzsch STA equipment used in this dissertation was calibrated for temperature and sensitivity using several melting standard reference materials. All experiments were performed in an argon atmosphere and the crucibles used were platinum-rhodium with alumina liners. An Oxygen Trapping System OTS, composed of a getter ring placed near the crucibles, was used in some instances to trap any residual oxygen in the instrument. Controlled cooling was achieved by injecting gaseous nitrogen into the system, when required. For all DSC experiments, an empty-crucible run was performed before the measurement of a sample to obtain an instrument baseline. This baseline was later subtracted from the DSC results obtained for the sample to get the heat flow signal from the sample alone. The enthalpy of any first-order transformation, in J/g, Proteus® software from the area under th

e DSC peak. Errors in temperature and en
e DSC peak. Errors in temperature and enthalpy of transformation determined from DSC experiments were evaluated by measuring an aluminum standard through the melting transition, and comparing the results to theoretical values. An error of ±0.5% and ±5% was determined for temperature and enthalpy, respectively. Isothermal DSC measurement performed in this dissertation were used to monitor nucleation and growth kinetics of the A1 transformation in FePd. The fraction transformed () into the L1 phase can be considered to be proportional to the amount of heat evolved, and thus it was determined by integrating the DSC peak that represents the transition: 87 ( 32 ) Where: = Fraction transformed, varying from 0 to 1 (dimensionless) = initial time of phase nucleation, after any initial incubation period (s) = final time, at which the transformation is complete (s) = time (s) h(t) = Heat flow during the transformation (mW/mg) The JMAK model (presented in Section 2.2.2.2) was used for the analysis of the DSC isothermal data. A graph of ln[ln(1/1-)] versus ln() was fit to a linear regression, and the Avrami exponent was extracted from the slope of the linear fit, while the rate constant was obtained from the intercept. Additionally, and considering a thermally-activated process, the rate constant follows an Arrhenius type behavior, from which the activation energy of the transformation ( 33 ) Where: = pre-e

xponential factor (dimensionless) = Act
xponential factor (dimensionless) = Activation energy (J/mol) = gas constant (J/mol·K) = Temperature at which the measurement was made (K) RESULTS AND DISCUSSION This dissertation reports on the effect of intrinsic and extrinsic modification on chemical ordering transformations in ferrous systems with the main goal of providing relevant information that can guide the laboratory synthesis of L1 FeNi, an attractive material for permanent magnet applications. To that end, FePd was chosen as a model system. The effect of () intrinsic modification through ternary alloying, and (modification through severe plastic deformation, on the stas well as on the thermodynamic and kinetic aspects of the A1 phase transformation, were studied in Aim 1 and Aim 2, respectively. Results pertaining to these Aims are presented in Sections 4.1 and 4.2. Furthermore, and considering the conclusions of Aim 1 and 2, ternary modification and severe plastic deformation were employed in the FeNi system to better understand aspects of L1 phase formation (Aim 3). The results obtained from experiments relevant to this Aim are presented in Aim 1: Understanding the effects of intrinsic modification, character, structure and magnetic This Section provides results obtained in fulfillment of Aim 1 of this dissertation. For organization and clarity, this Section has been divided into five parts: Section 4.1.1 provides a brief introduction of the s

tudies performed here on ternary FePdM (
tudies performed here on ternary FePdM (M = Ni or Cu) alloys and a justification of the ternary element selection; Sespecific details of sample synthesis, processing and characterization; Section 4.1.3 provides structural, magnetic and calorimetric results of the FePd, FePdNi and FePdCu alloys studied; Section 4.1.4 presents an analysis and discussion of the results while As described in Section 2.1.4.1 substitutional ternary alloying additions are frequently used in L1-forming ferrous systems to tune their magnetic properties, or to modify thermodynamic and kinetic parameters of the order-disorder transformation to meet specific requirements. Nonetheless, an understanding of the relations between the nature and the quantity of the ternary element on the structure, magnetism, and chemical ordering characteristics of L1-forming ferrous systems is underdeveloped, and a model with predictive capabilities is not yet a reality. To contribute useful information that can complement previous studies and ultimately provide a basis to advance this understanding, studies of the effects of Ni and Cu ternary alloying additions on the phase transformation character, structure and magnetic properties of the model FePd system were carried out in this dissertation. The choice of ternary Ni additions was based on the fact that the FeNi system exhibits an L1 phase near the equiatomic composition that is very attractive for perman

ent magnet applications. Thus, incorpora
ent magnet applications. Thus, incorporation of Ni in the formula Fe was chosen to allow gradual approach to the L1 FeNi phase, and to simultaneously determine the effects of this addition on the structure, magnetism and phase formation character on the FePd compound. Additions of Cuwere selected on the basis of the similarity of Cu and Ni; both are 3-transition metals corresponding to adjacent groups in the periodic table, they have similar atomic size, similar electronegativities, and both crystallize into an fcc structure of comparable lattice parameters. However, as Ni is a ferromagnetic metal at room temperature and Cu is not, a comparison of ternary FePd formed with Ni and Cu additions is anticipated to provide relevant information to better understand the effects of magnetism on the chemical ordering transformation of FePd. FePdM sample synthesis and characterization Alloys with the nominal starting composition Fe (M = Ni or Cu, = 0, 3, 5, 7 at%) were synthesized from high-purity elemental granules of Pd (99.9%), Fe (99.98%), Ni (99.999%) and Cu (99.98%) by arc-melting. Disks (~3 mm in diameter) were sliced out of the arc-melted charges for further processing and analyses. Composition and chemical homogeneity were verified by standardless SEM/EDS in selected alloy slices that were polished with a final grit size of 0.5 m. To induce L1chemical ordering, FePdM samples were encapsulated in evacuated (1 x

10 Torr) silica tubes and annealed for 1
10 Torr) silica tubes and annealed for 100 h at 773 ± 5 K. The annealing conditions were selected according to the time-temperature-transformation () diagram for FePd described in Section 2.1.2.1, which indicates a high rate of chemical ordering in the vicinity of 773 K, with more than 90% of the transformation into the L1 phase complete for annealing times of 100 h. The crystal structure of the FePdM samples both in the as-arc-melted and as-annealed states was examined by XRD. The degree of chemical order as characterized by parameter was determined from the integrated intensity of the () superlattice ) fundamental Bragg peaks (as described in Section 3.3.2), corrected for effects of finite sample size. The finor as a function of the X-ray incident angle was determined as described in Appendiy, in circumstances of multiphase coexistence, the volume fraction of a given phase was determined from XRD data using a relation of the intensities of the main Bragg reflections of each phase The magnetic character of as-arc-melted and as-annealed FePdM samples was investigated as a function of magnetic field using vibrating sample magnetometry (VSM, Quantum Design Versalab). In-plane hysteresis loops of the disk-shaped samples were obtained at room temperature with a maximum applied field of 30 kOe. Demagnetization corrections for the specific sample geometries were determined from tabulated data [133], as described in

Section 3.3.3. For samples in which the
Section 3.3.3. For samples in which the maximum applied field was not sufficient to achieve saturation, was estimated graphically from the plot of , as described by the Law of Approach to Saturation for polycrystalline materials described in Section 3.3.3. Structural and magnetic phase transformations were studied for both as-arc-melted and as-annealed FePdM alloys using constant heating rate calorimetry. The reversibility of any transformation observed during the first DSC heating scan was confirmed by cooling at a rate of 20 K/min and performing a second heating run on the same sample. The enthalpy of any first-order phase transformation, described in Section 3.3.4), was transformed to kJ/mol using the molecular weight of the corresponding composition. In cases of multiphase coexistence, the enthalpy in J/g was first corrected to account for the mass ng the phase transformation. Comparison of the Curie transition for chemically-modified Fe50-x (M = Ni or Cu) alloys, determined from calorimetry, was made on the basis of the average weighted valence band electron concentration. This valence electron concentration criterion has been commonly used in the study of 3transition metal substitutional alloys to relate the effect of composition on a variety of physical parameters such as Curie temperature and average saturation magnetization [173,174]. This criterion was used here to examine the dependence of on com

position for FePd-based ternary alloys.
position for FePd-based ternary alloys. Calculation of the average weighted valence band electron concentration for the ternary FePdM compositions studied used the following equation: ( 34 ) Where are the atomic fractions of the elements Fe, Pd and M and , are the valence electrons of Fe, Pd and M elements. All elements involved are transition metals, are usually considered as the number of electrons in the outermost The results presented in this Section have been organized such that those from the binary FePd system are described first, followed by those from FePd modified with ternary Ni or Cu additions. For each system, structural, magnetic and calorimetric results are presented for both the as-arc-melted and annealed states. The average composition of the binary FePd alloy was confirmed to be 47.552.5 (at %) with a standard deviation of 1 at%. The small standard deviation is evidence of good chemical homogeneity. The XRD data from the FePd sample in the as-arc-melted state exhibits no crystalline phases other than the fcc chemically disordered phase. Figure 34a presents the X-ray diffractogram obtained for this as-arc-melted FePd sample, where it can be seen that all Bragg reflections for this sample match those reported for an equiatomic fcc FePd alloy [175]. Upon annealing, the alloy transforms into a single L1 phase, as evidenced by the X-ray annealed sample which is in agreement with that reporte

d for equiatomic L1 FePd [48] (Figure 34
d for equiatomic L1 FePd [48] (Figure 34b). The presence of L1-type superstructure peaks in this sample ((), (), and () among others) indicates achievement of chemical ordering, while the splitting of selected fundamental Bragg reflections ((), ()) evidences the reduction from cubic to tetragonal symmetry. The lattice parameters obtained for the binary as-arc-melted fcc-type FePd ( = 3.820 ± 8 Å) and annealed L1type FePd ( = 3.697 ± 1 Å) are in broad agreement with those reported in literature for a similar composition [26,49], and indicate a relative reduction in unit cell volume of 1% upon the fcc transformation. Peak broadening analysis of the XRD data through a modified Williamson-Hall plot (Figure 35) reveals that the dimension of coherently diffracting domains for the as-arc-melted fcc FePd phase is 43 ± 10 nm, while that for the annealed L1 FePd phase is 57 ± 14 nm. The estimated microstrain in these samples is 0.25 ± 0.04% and 0.19 ± 0.04% for fcc and L1 FePd, respectively. The parameter of the annealed L1–type FePd of this study was determined from X-ray The room-temperature magnetization of the fcc (as-arc-melted) FePd phase increases rapidly with increased applied field (Figure 36), reaching a saturation value of = 1066 ± 4 emu/cc, in good agreement with that reported in the literature for bulk fcc [10]. The coercivity and remanence for this sample are negligible, with values 5 Oe and 40 emu/cc.

In contrast, the magnetization of the an
In contrast, the magnetization of the annealed L1 FePd slowly increases with applied field without reaching saturation at 30 kOe, indicative of increased magnetocrystalline anisotropy relative to the fcc FePd phase. The extrapolated value, as determined graphically from the Law of Approach to Saturation (described in Section 3.3.3), is higher than that of the fcc counterpart ( = 1180 ± 9 emu/cc), as are the determined coercivity and remanence parameters with values of 94 Oe and Figure 34. X-ray diffractograms obtained for binary FePd in the a) as-arc-melted state, and b) annealed state. Reference patterns of fcc FePd (JCPDS 04-rence patterns of fcc FePd (JCPDS 04-) and L10 FePd (JCPDS 03-065-9971 FePd (JCPDS 03-065-9971 ) are included for comparison. 95 Figure 35. Modified Williamson-Hall plot for FePd in its as-arc-melted (blue, open circles) and annealed (red, closed squares) state. The broken lines represent the linear fit to the data, from which the intercept and slope are extracted to estimate the crystallite size and microstrain. Figure 36. Magnetization as a function of applied field measured at room temperature for FePd in its as-arc-melted (blue) and annealed (red) states. Examination of the DSC data that provi heat flow as a function of increasing temperature reveals a single calorimetric event of a second-order character for FePd in its fcc state (as-arc-melted), as shown in Figure 37. This

feature is associated with the Curie tem
feature is associated with the Curie temperature, determined as = 724 K. In contrast, the annealed L1-type FePd sample shows multiple thermal features (Figure 37): a Curie transition at = 737 K is followed by a broad irreversible endothermic event in the range 770 K 900 K. Next, a first-order sharp irreversible endothermic transition in the range 900 K is observed, with = 944 K. A combination of both endothermic events (broad and sharp) is attributed to the L1fcc order-disorder phase transformation, with the area under the curve representing the enthalpy of transformation = 24.5 J/g (2.0 kJ/mol) and with onset = 944 K indicative of the order-disorder temperature . The transformation is irreversible, and a subsequent DSC heating scan scan in Figure 37) reveals information of a “recovered” phase. A single second-order calorimetric event is observed for this recovered phase, associated with its Figure 37. Differential scanning calorimetry results at 20 K/min on FePd in the as-arc-melted (blue, top) and annealed (red, bottom) states. For the annealed sample, consecutive heating scans are labeled “1FePd modified with Ni ternary alloying additions The average compositions of the ternary FePdNi alloys are presented in Table 4. For simplicity, from this point onward the alloys will be referred to by their determined Ni content as = 3.5, 5.2 and 7.5. The low standard deviation of the measurements emical homogen

eity in these alloys. ns (at%) of ternar
eity in these alloys. ns (at%) of ternary FePdNi samples. Fe Pd Ni Nominal 50 47 3 Measured 49.3 (0.5) 47.2 (0.4) 3.5 (0.3) Nominal 50 45 5 Measured 47.5 (0.8) 47.3 (0.7) 5.2 (0.2) Nominal 50 43 7 Measured 48.2 (0.4) 44.3 (0.4) 7.5 (0.1) The XRD data obtained from ternary FePdNi alloys in the as-arc-melted state exhibit no crystalline phases other than the metastablly-retained fcc chemically disordered phase, as confirmed by laboratory XRD (Figure 38a). Upon annealing, the = 3.5 sample transforms into a nominally single-phase L1 structure, as evidenced by the appearance of superstructure peaks and the splitting of the fundamental (), () Bragg reflections (Figure 38b). Annealed samples with a Ni content 5.2 on the other hand, show a multiphase character. Superstructure peaks are observed for these samples, signaling the presence of chemical ordering and thus the existence of an L1phase. This conclusion is corroborated by the splitting of the fundamental (), ( a phase of tetragonal symmetry. However, these split peaks are accompanied by additional Bragg reflections which indicate the existence of additional phase(s) (see Figure 39 for a magnified view of ) diffraction peak). For the annealed = 5.2 sample, these additional reflections are attributed to the presence of an fcc-based phase with a lattice parameter that is the same, within experimental error, as that of the fcc phase in the as-arc-melted sampl

e. Thus, this fcc phase has been tentati
e. Thus, this fcc phase has been tentatively attributed to a remnant fcc parent phase. On the other hand, the = 7.5 annealed sample shows two sets of additional Bragg reflections, corresponding to two fcc-based phases of different composition. One of these fcc phases in the annealed = 7.5 sample has the same lattice parameter, within experimental error, as that of the fcc as-arc-melted = 7.5 phase, and is thus tentatively attributed to a remnant fcc parent phase. The other fcc phase observed has a much smaller lattice parameter than that of the parent fcc phase (relative difference in unit cell volume of 1.3%), and is thus considered to be a “new” fcc phase. The L1 phase volume fractions in the = 5.2 and 7.5 annealed samples were determined as 95% and 20%, respectively, while the fcc phase volume fractions in the = 7.5 annealed sample were estimated as 99 Figure 38. X-ray diffractograms obtained for ternary FePdNi samples in their a) as-arc-melted state, and b) annealed state. Samples are identified by their determined Ni content. Diffractograms obtained for binary as-arc-melted and annealed FePd ( = 0) are included for comparison, as well as FePd (JCPDS 04-003-5130 (JCPDS 04-003-5130 ) and L1Figure 39. Multi-peak structures in the vicinity of the () diffraction peak for a) = 5.2 sample and b) annealed = 7.5 sample. Peaks labeled with “*” are attributed to fcc phase(s). Additions of Ni are found to

have a notable effect on the structural
have a notable effect on the structural characteristics of fcc-type and L1-type FePd. Figure 40 shows trends observed for changes in the unit cell volumes, ratios, parameters, crystallite sizes () and microstrain (increased Ni content. Values for the binary unmodified FePd sample ( = 0), discussed in Section 4.1.3.1, are included for reference. The unit cell volumes of the fcc (as-arc-melted) and the L1 (annealed) phases decrease with increased Ni content. The degree of tetragonality of the L1 phase found in the annealed samples, as represented by the ratio, remains relatively stable with Ni incorporation at a value of ~0.96, while the LROparameter was observed to decrease upon addition of 3.5 at% Ni. The parameter determined for samples with higher Ni content ( 5.2) has a large associated error, attributed to the deconvolution process of the individual L1 and fcc reflections in the multi-peaks observed. Therefore, these values are omitted from Figure 40. The calculated crystallite sizes for the fcc phase found in as-arc-melted samples decrease with increased Ni concentration, from 43 nm for the binary FePd sample to 20 nm for the FePdNi composition. Crystallite sizes of all FePdNi samples increase systematically upon annealing and formation of the L1 phase. The calculated microstrain values are moderate, in the range 0.10% 0.25%, and do not follow a systematic variation with Ni additions. For a given compos

ition, the strain manifest in the fcc (a
ition, the strain manifest in the fcc (as-arc-melted) FePdNi phase is the same, within experimental error, as that characterizing the tetragonal 101 Figure 40. Variation with Ni content of unit cell volume, ratio, parameter, crystallite size () and microstrain () of the fcc (as-arc-melted) and L1(annealed) phases. Results for binary FePd ( = 0) have been included for comparison. Lines are drawn to guide the eye. The room-temperature magnetization behavior of the fcc (as-arc-melted) ternary FePdNi samples is very similar to that of the binary fcc (as-arc-melted) FePd sample, increasing rapidly towards saturation with increased applied field (Figure 41a). This same behavior is also displayed by the annealed = 7.5 FePdNi composition (Figure 41b). Coercivity and remanence parameters for these soft ferromagnetic FePdNi samples are negligible, with values and 60 emu/cc. In contrast, the magnetization behavior of annealed 5.2 FePdNi samples slowly increases with applied field, without reaching saturation at 30 kOe, similar to the behavior found for the annealed binary FePd sample. The coercivity and remanence parameters of these annealed FePdNi (samples are higher than those of their fcc (as-arc-melted) counterparts, with values 157 emu/cc. Saturation magnetization values determined graphically from the Law of Approach to Saturation for FePdNi samples in their annealed state are higher than those of the corr

esponding as-arc-melted state; a relativ
esponding as-arc-melted state; a relative increase of 7 to 10% is seen for 5.2 samples upon annealing, while a modest 2% increase in is Figure 41. Magnetization as a function of applied field measured at room temperature for ternary FePdNi samples in their a) as-arc-melted, and b) annealed states. Samples are identified by their determined Ni content. ) results obtained for binary as-arc-melted and annealed FePd (are included for comparison. 103 with Ni content for as-arc-melted and annealed ternary FePdNi samples. Results for binary FePd ( = 0) are included for reference. Lines are drawn to guide the eye. Note that the annealed sample does not follow the trend line; the error bars for this sample are smaller than the symbol. The DSC data measured from all as-arc-melted ternary FePdNi samples display a single calorimetric event of a second-order character, as shown in Figure 43, associated with the Curie transition of the fcc phase. Annealed FePdNi alloys with Ni content 5.2 show multiple thermal features in the DSC data, similar to those observed for annealed binary FePd (Section 4.1.3.1); a second-order event, associated with the Curie transition of the L1 FePdNi phase, is followed at higher temperatures by a broad irreversible endothermic event starting at = 770 K that precedes the endothermic first-fcc phase transformation which occurs at a different temperature for each sample. The transformation

is irreversible, and a second heating sc
is irreversible, and a second heating scan carried out on these samples reveals information of a “recovered” phase, showing only a Curie transition. For = 7.5 annealed sample, only a Curie transition is detected by calorimetry, despite the fact that the XRD data evidenced approximately 20 vol% of the L1 phase in this sample; no measurable enthalpy associated with the chemical order-disorder phase transformation in the FePdNi Figure 43. Differential scanning calorimetry results at 20 K/min of ternary FePdNi samples in their as-arc-melted and annealed states. Samples are identified by their determined Ni content. DSC results for binary FePd ( = 0) are included for reference. For all annealed samples, consecutive heating FePd modified with Cu ternary alloying additions The average composition of the ternary FePdCu alloys are presented in Table 5. For simplicity, from this point onward the alloys will be referred to by their determined Cu content as = 4.3, 5.8 and 7.9. The low standard deviation of the measurements ()f their chemical homogeneity. ns (at%) of ternary FePdCu samples. Fe Pd Cu Nominal 50 47 3 Measured 48.9 (0.6) 46.8 (0.7) 4.3 (0.3) Nominal 50 45 5 Measured 49.0 (0.6) 45.2 (0.3) 5.8 (0.7) Nominal 50 43 7 Measured 49.2 (0.6) 43.0 (0.6) 7.9 (0.6) As-arc-melted FePdCu alloys consist of a single fcc chemically disordered phase, as confirmed through laboratory XRD (Figure 44a). Upon annealing, th

e presence of superstructure peaks as we
e presence of superstructure peaks as well as the evident splitting of some fundamental Bragg attainment of the L1 structure in these samples. An additional Bragg 45° is observed in all annealed FePdCu samples, which coincides with the most intense Bragg peak of a -Fe (A2-type Fe) phase. Thus, annealed FePdCu samples are composed of a phase mixture -Fe + L1 FePdCu. The volume fraction of the phase was estimated from a relation of the integrated intensities of the main Bragg peaks of both phases (see Section 3.3.2 for details) as 96, 92, and 87% for the = 4.3, 5.8 and 7.9 compositions, respectively. 106 Figure 44. X-ray diffractograms obtained for FePdCu samples in their a) as-arc-melted state, and b) annealed state. Samples are identified by their determined Cu content. Diffractograms obtained for binary as-arc-melted and annealed FePd ( = 0) are included for comparison, as well as reference patterns for fcc FePd (JCPDS 04-003-5130 FePd (JCPDS 04-003-5130 ), L1 FePd FePd ), and A2 Fe (JCPDS 03-065-9971 [176]). The additions of Cu to FePd impact the structural characteristics of fcc-type and -type FePd, as observed in Figure 45, where changes in the unit cell volumes, parameters, crystallite sizes () and microstrain () are shown as a function of increased Cu content. Values of these parameters for the binary FePd sample ( = 0), discussed in Section 4.1.3.1, are included for reference. The unit cell volum

e of the fcc (as-arc-melted) phase decre
e of the fcc (as-arc-melted) phase decreases slightly with increased Cu content, while within error it remains constant for the L1 (annealed) phase. The tetragonality of the L1 (annealed) phase is greatly affected by Cu incorporation, with the c/a ratio decreasing sharply with increasing Cu additions. The parameter of L1 FePdCu is relatively constant with varying Cu content, with an average value of 0.83. There is no significant change in both the crystallite size and microstrain with Cu incorporation for both the fcc (as-arc-melted) (annealed) phases. In all studied FePdCu alloys, the crystallite size of the L1(annealed) phase is consistently higher than that of its fcc (as-arc-melted) counterpart, while microstrain is consistently lower. Figure 45. Variation with Cu content of unit cell volume, ratio, parameter, crystallite size () and microstrain () of the fcc (as-arc-melted) and L1(annealed) phases. Results for binary FePd ( = 0) have been included for comparison. Lines are drawn to guide the eye. The room-temperature magnetization behavior of the fcc (as-arc-melted) ternary FePdCu samples increases rapidly to saturation with increased applied field (Figure 46a). These sample show negligible coercivity and remanence values, and In contrast, the magnetization of annealed samples increases slowly with increased applied field, indicative of an enhancement in magnetocrystalline anisotropy upon anneal

ing (Figure 46b). Magnetic saturation is
ing (Figure 46b). Magnetic saturation is not reached at the maximum applied field of 30 kOe, but extrapolated values are higher than those of the fcc (as-arc-melted) counterparts (Figure 47). The coercivity and remanence of FePdCu samples increases upon annealing and formation of the L1 phase, achieving values u/cc. Figure 46. Magnetization as a function of applied field measured at room temperature for ternary FePdCu samples in their a) as-arc-melted and b) annealed states. Samples are identified by their determined Cu content. ) results obtained for binary as-arc-melted and annealed FePd (are included for comparison. 109 with Cu content for as-arc-melted and annealed ternary FePdCu samples. Results for binary FePd ( = 0) are included for reference. Lines are drawn to guide the eye. DSC data obtained from the as-arc-melted ternary FePdCu samples display a single calorimetric event of a second-order character in DSC results, associated with the Curie transition of the fcc phase (Figure 48). On the other hand, annealed FePdCu samples exhibit multiple thermal features in DSC; a Curie transition is observed first, associated with the L1 phase, and is followed by a broad irreversible endothermic event in the range 770 K 925 K. Next, a first-order endothermic peak with a multi-stage character is observed in the range 925 K 1100 K. The multiple stages of this thermal event have onset temperatures that vary wi

th composition, as seen in Figure 49. Th
th composition, as seen in Figure 49. The combination of all observed endothermic events is associated with a eutectoid fcc phase transformation. The transformation is irreversible, and a second heating scan on these FePdCu samples reveals a second-order thermawith a Curie transition of a “recovered” phase. For the = 7.9 annealed FePdCu sample, the second heating scan also reveals two small features at = 933 and 980 K, which 110 Figure 48. Differential scanning calorimetry results at 20 K/min of ternary FePdCu samples in their as-arc-melted and annealed states. Samples are identified by their determined Cu content. DSC results for binary FePd ( = 0) are included for reference. For all annealed samples, consecutive heating 111 Figure 49. Individual onset temperatures identified for the two endothermic features that conform the first-order peak observed in DSC for ternary FePdCu annealed samples. For the 5.8 annealed FePdCu samples, for both features is clearly identifiable, while for the = 7.9 FePdCu annealed sample deconvolution into two assymetric peaks was necessary to get an estimate of onsetIn this Section, the effects of ternary alloying additions of Ni or Cu to near-equiatomic FePd are discussed. The particular aspects that are examined are categorized in four different subsections: first, a discussion of the effects of ternary alloying additions of Ni/Cu on the phase equilibria of the FePd system

at 773 K is presented. The isothermal 7
at 773 K is presented. The isothermal 773 K sections of the ternary Fe-Pd-Ni and Fe-Pd-Cu phase diagrams near the equiatomic FePd composition are obtained from the experime this dissertation. Next, the effects of ternary alloying additions of Ni/Cu on the structure of fcc and L1FePd is shown, demonstrating contrasting effects of both types of additions on the characteristics of the L1 phase. The effects of ternary alloying additions of Ni/Cu on the magnetic properties of fcc and L1 FePd are analyzed subsequently, demonstrating that the Curie transition varies systematically with the valence electron concentration of the alloy. Finally, the effects of Ni or Cu substitution on the phase transformation behavior of FePd is discussed, demonstrating that the A1 transition is very sensitive to ternary element incorporation. Effects of ternary additions on phase equilibria at 773 K in near-In this section, the structural data obtained for all studied FePdM compositions at 773 K is analyzed in the context of phase s obtained in this dissertation, together with those presented by other authors, are used to generate the 773 K isothermal sections of the ternary Fe-Pd-Ni and Fe-Pd-Cu systems near the compositions. It is demonstrated that increased additions of Ni into FePd quickly destabilize the L1 phase, while Cu has a much moderate effect. 47.552.5 binary alloy at 773 K for 100 h produces a single L1phase (Figure 34),

as expected at equilibrium for this com
as expected at equilibrium for this composition according to the Fe-Pd phase diagram, Section 2.1.2.1. The equilibrium L1 structure is obtained with an imperfect chemical order ( = 0.72), consistent with the off-stoichiometry of the mperature (see Section 2.1.3.1). The phases detected in all annealed (773 K, 100 h) ternary FePdNi samples (Figure 38) are in qualitative agreement with those presented in the calculated ( = 823 K) Fe-Pd-Ni ternary phase diagram of Horiuchi et al. [110]. As seen in Section 2.1.4.1, Horiuchi's phase diagram indicates the existence of a single L1 phase centered around the Fe composition next to a two-phase “disorder” + L1 region, Figure 50a. The solubility limit of Ni in the single L1 phase at a fixed Fe content (~50 at%) is provided by Horiuchi et al. as ~5 at%. Consistent with this scenario, in the present study phase was detected in annealed 3.5 FePdNi samples. For the FePdNi annealed alloy, the fcc phase transformation did not go to completion, an a small volume fraction (~5 vol%) of the parent fcc phase remained. It is anticipated that this particular allsform into a single L1 phase if sufficient time is allowed to achieve equilibrium. The observed phases present in the FePdNi annealed alloy are also consistent with Horiuchi et al.'s proposed diagram, for which a two-phase mixture is expected to exist in equilibrium; these two phases correspond to the observed “new” fcc and L1

phases, while the presence of a remnant
phases, while the presence of a remnant fcc parent phase indicates that the transformation also did not go to completion in this sample. This apparently good agreement between the current reet al. should be considered with caution as () the calculations of Horiuchi et al.[110] predict = 800 K for binary L1 FeNi, significantly higher that the experimentally determined value of 593 K [16,56,65] and ( FePd boundary in et al.'s calculated ternary phase diagram does not agree with the published Fe-Pd binary phase diagram [26]. Nonetheless, Horiuchi et al phase boundary near the 50 at% Fe composition has been used to estimate the Ni concentration of the phases present in the = 7.5 FePdNi annealed alloy. By invoking assumptions that volume percent is equivalent to mass fraction, and that the L1 phase and the parent disordered fcc alloy possess the same Fe concentration (48.2 at%), the Ni concentrations of the phases present in the = 7.5 FePdNi annealed alloy are estimated by conservation of mass as = 6 at% (L1 phase) and as = 8 at% (new fcc phase), (Figure 50b). This information allows determination of a very narrow fcc + L1 two-phase region for the FePdNi system in the composition range 6 at% or 48.2 at% Fe at 773 K. Figure 50. a) Fe-Pd-Ni ternary phase diagram at 823 K proposed by Horiuchi et alet al, along with the alloy compositions of the present study. b) Estimated compositions of the multiple phases pre

sent in the = 7.5 annealed sample: pare
sent in the = 7.5 annealed sample: parent fcc phase (blue), product L1 phase (red), product new fcc phase (green). Horiuchi et al.'s L1 boundary is left for reference as a dashed line since it was used to estimate the compositions. Using the estimated compositions for the two-phase fcc + L1 region in the FePdNi system, along with phase equilibria information at 773 K obtained from the binary Fe-Ni [177], Ni-Pd [178], and Fe-Pd [26] phase diagrams, an updated isothermal section of the ternary Fe-Pd-Ni phase diagram is proposed in Figure 51. It is noted that the phase boundaries presented for single-phase regions other than L1 are arbitrary, and are included for illustration purposes only. Assignment of a single-phase L1 region extending across the full composition range is made based on the complete mutual solubility of Ni and Pd in the fcc phase in the NiPd binary system. Subjecting the = 5.2 = 7.5 FePdNi alloys to longer annealing times should prove useful in the determination of the L1 and the fcc + L1 phase boundaries with greater certainty. In the proposed 773 K isothermal section of the Fe-Pd-Ni system, an fcc phase is shown to be more stable than an L1 phase for Fe alloys with Ni content higher than Figure 51. 773 K isothermal section proposed for the updated Fe-Pd-Ni ternary phase diagram ( represents the body centered cubic crystal structure represents an fcc phase). L1 and fcc + L1 phase boundari

es have been selected to match experimen
es have been selected to match experimental observations, while phase boundaries displayed for other regions are arbitrary. Analysis of the structural results obtained for the FePdCu alloys annealed at 773 K for 100 h (Figure 44) reveals achievement of an phase mixture upon annealing. The phase mixture obtained for these compositions at 773 K is in agreement with results obtained by Naganuma et al. [107] in Fe nanoparticles � 45 at% fabricated by electron-beam evaporation followed by post-deposition annealing at 823 K for 1 h. A compilation of results obtained by different authors on FePdCu thin films or nanoparticles, with a composition similar to the alloys studied in this dissertation and exposed to annealing conditions similar to those used here, is shown in Figure 52. Data points in this figure are represented by open symbols whenever an phase mixture was detected upon annealing, and with closed symbols when a phase was obtained. From this information, and assuming that equilibrium conditions were achieved in the reference studies, an equilibrium phase field is proposed at 773 K for ternary FePdCu alloys with low Cu content (at%) and Fe 48 at%, while an equilibrium single-phase L1 region is proposed for the same Cu contents but lower atomic percent Fe. Furthermore, considering the equilibrium phases reported at this temperature for the Fe-Pd [26], Fe-Cu [32], and Cu-Pd [179] binary systems, a

n isothermal section of the ternary Fe-P
n isothermal section of the ternary Fe-Pd-Cu is proposed in Figure 53. It is noted that the phase boundaries presented in this ternary phase diagram, other than that phase, are arbitrary and are included for illustration purposes only. Figure 52. FePdCu ternary diagram with compositions from this study (annealed at 773 K) and those reported in literature for a similar annealing temperature (773 – 873 K) K) . In gray text, the equilibrium phases expected for binary FePd are shown. The data points are color/symbol-coded to indicate the phases detected in these annealed alloys; blue (closed symbols) indicates a single L1 phase and green (open symbols) indicates an 117 Figure 53. 773 K isothermal section proposed for the ternary Fe-Pd-Cu phase diagram ( represents the body centered cubic crystal structure, represents an fcc phase, represents a B2 phase). The L1 phase boundary has been drawn to match experimental observations in this work and results reported by other authors s . Phase FePd In this Section, the structural character of the A1 and L1 phases present in type FePd phase are examined first, followed by results from the L1-type phase. It is shown that Ni and Cu additions affect the A1 lattice in the same way, but have lattice. In the as-arc-melted state, a single fcc phase was obtained for all studied compositions (Figure 34, Figure 38, and Figure 44), confirming that the cooling rate after arc-melti

ng is sufficiently fast to metastably re
ng is sufficiently fast to metastably retain the high-temperature chemically disordered phase to room temperature. The lattice parameters of this fcc phase vary with ternary additions of Ni or Cu in a similar way (Figure 40, Figure 45): both additions decrease the lattice parameter with increased ternary element substitution at a similar rate, an produce an overall decrease in unit cell volume. These results indicate that, in principle, Ni and Cu behave in a similar manner in the FePd lattice, with substitution of Pd atoms for the smaller Ni/Cu atoms resulting in an overall contraction of the lattice. The structural attributes of the L1 phase obtained in all annealed FePdM samples varies with the amount and identity of the ternary element M. Increased additions of Ni to FePd contract the unit cell volume of the L1 phase, with no major impact on the tetragonality ( ratio) (Figure 40). Increased Cu content in FePdCu alloys does not modify the unit cell volume of the L1 phase formed, but does drastically modify the tetragonal distortion, as represented by smaller ratios (Figure 45). There is no clear trend in the evolution of the degree of chemical order in L1 FePdM, as represented by parameter, with the amount and identity of the ternary element M. on the magnetic properties of near- FePd In this section, the magnetic properties of the A1 and L1 phases present in FePdM (M = Ni or Cu) alloys as a function of M

content will be anA1-type FePd phase are
content will be anA1-type FePd phase are examined first, followed by results on the L1 phase. It is shown that Ni and Cu additions have contrasting effects on the Curie temperature of both phases, and that generalized trends can be made when representing the alloy composition The fcc phase present in all as-arc-melted FePdM samples studied has a very soft ferromagnetic behavior, as expected for a high-symmetry cubic phase. In contrast, annealed alloys with the L1 phase show a very slow approach to saturation, indicative of their higher magnetocrystalline anisotropy (Figure 36, Figure 41, Figure 46). For the = 7.5 FePdNi annealed sample, the magnetic behavior is dominated by the majority fcc phases. Additions of Ni into fcc and L1–type FePd have a non-systematic effect on the , while increasing the Figure 54). Increased additions of Cu into fcc and -type FePd decrease both intrinsic magnetic properties and (Figure 47, Figure Figure 54. Evolution of the Curie temperature of L1 and fcc phases in a) FePdNi and b) FePdCu alloys as a function of Ni/Cu content. In (a), for the FePdNi annealed sample, note is plotted separately from the trend line presented for annealed L1 samples, due to the low volume fraction (~20%) of L1 phase in this sample. The effects of Cu and Ni additions on of the fcc and L1 phases are presented and compared in terms of the average valence the alloys. A plot vs. average valence electro

n per atom (Figure 55), reveals that of
n per atom (Figure 55), reveals that of both fcc and ternary FePdM alloys decreases systematically with increased valence electron concentration. The behavior for each FePdM phase echoes the behavior of for binary FePd with varying Fe-content, represented as a solid line in Figure 55. Thus it is concluded that ternary 3-element additions can be chosen based on their electronic configuration to tune the Curie temperature of near-equiatomic fcc and L1 FePdM ternary alloy. Figure 55. Variation of the Curie temperature of a) fcc phase, and b) L1 phase, in near-equiatomic FePd-based alloys versus alloy composition, represented by the valence electrons per atom. The variation in the Curie temperature of fcc and L1 phases in binary FePd as a function of composition, are included for comparison on . Finally, it should be noted that once all studied alloys undergo a phase transformation upon heating in DSC (L1fcc for FePd and FePdNi, and fcc for FePdCu), the high temperature fcc phase is retained to room temperature upon cooling. The Curie temperature identified for this recovered fcc phase provides a means to relate its structural state to that of the fcc as-arc-melted phase. In the binary FePd and in ternary FePdCu samples, this recovered fcc phase must be structurally equivalent to the as-arc-melted fcc phase, as demonstrated by identical of both phases, Figure 54. The Curie temperature of the recovered phase

in the ternary FePdNi samples is higher
in the ternary FePdNi samples is higher than that of the as-arc-melted fcc phase (Figure 54), indicative of structural differences between them. It is proposed that retention of a small degree of short-range order occurs in these FePdNi alloys after these samples undergo a chemical disordering L1near-equiatomic FePd In this Section, the influence of ternary alloying additions on the A1transformation in FePd is discussed. First, salient aspects of the L1A1 transformation in unmodified FePd are examined, including the occurrence of a two-step chemical disordering process. Then, the effects of Ni on the and on the enthalpy of the A1 transformation L10 of FePd is presented, and implications regarding the reverse transformation (A1) upon isothermal annealing are discussed. Last, the effects of Cu additions on the eutectoid + L1A1 transformation of FePd are evaluated. Overall, it is determined that Ni additions drastically reduce the and the L10of the chemical order-disorder transformation, resulting in a lower driving force that is consistent with the observed sluggish phase transformation kinetics in this system. Data collected during isochronal testing of binary L1 FePd samples in the DSC exhibits a sharp endothermic peak attributed to a chemical disordering L1fcc transformation of first-order character (Figure 37). The onset temperature of this transition (944 K), taken as an indication of , is very close to t

he equilibrium expected for an alloy of
he equilibrium expected for an alloy of this composition (983 K) [26]. A broad endothermic event is observed preceding this chemical order-disorder transformation, with an onset temperature of ~770 K. While broad endothermic signals that precede a chemical order-disorder transformation have been reported for thermo resistometry [184] experiments conducted on chemically ordered CuAu samples, their origin remains unexplained. In this dissertation, this feature is attributed to atomic short-range motion below the temperature at which the nucleation and growth of the chemically-disordered fcc phase may be measured, to produce a readjustment of the parameter to equilibrium values. In this context, the total enthalpy of the L1fcc transformation is taken as the area under both endothermic evenfcc transformation in ternary FePdNi also exhibits a two-step character, as found in binary FePd, Figure 43. Incorporation of Ni into FePd of the order-disorder transformation fcc and systematically reduces the critical order-disorder temperature , Figure 56. The lower and lower fcc found in FePdNi alloys with increased Ni content is consistent with the sluggish kinetics observed phase formation: a lower produces a lower undercooling upon annealing at 773 K, which together with a decreased enthalpy of transformation reduces the driving force for the chemical ordering process in ternary FePdNi samples. Figure 56. Evolution

of the onset temperature and enthalpy o
of the onset temperature and enthalpy of the order-disorder fcc phase transformation in FePdNi as a function of Ni content. For = 7.5 FePdNi annealed sample, note there is no data point for , as no measurable enthalpy associated with the L1fcc transformation was fcc eutectoid transformation experienced by FePdCu alloys upon heating in DSC (Figure 48, Figure 49) reveals two convoluted endothermic peaks, with compositionally dependent onset temperatures and enthalpies of transformation. These endothermic peaks are attributed to the different stages of this eutectoid phase transition. While additions of Cu have only a small effect on the onset temperatures of the individual stages of the fcc eutectoid transformation, the overall enthalpy of the transformation is significantly increased, Figure 57. These results indicate that Cu additions have a pronounced effect on the driving force of this eutectoid transformation in the FePdCu system. The effect of Cu additions on the chemical order-disorder phase transformation (L1fcc) of FePd could not be established from these results. However, it has been reported that 7 at% addition of Cu into FePd decreases of the L1transformation by 50 K [107]. It was found in this dissertation that additions of 5.2 at% Ni into FePd caused a reduction of 94 K in , as shown in Figure 56. The apparent slower rate at which Cu additions decrease in comparison with that produced by Ni addit

ions may be related to the observed diff
ions may be related to the observed difference in the stability range of the L1 phase. It is anticipated that ternary element additions that increase can favor the L1 phase in the ternary compound. 124 Figure 57. Evolution of the onset temperatures and enthalpy of transformation as a function of Cu content in FePd. For binary FePd, the transformation is fcc. For ternary FePdCu the transformation is + L1fcc. Lines are drawn to guide the eye. Significance of work on composition-structure-property correlations in FePd The effects of intrinsic modification on the structure-property correlations as well as on the phase transformation character of FePd, have been examined here through study of samples formed with ternary alloying additions of Ni and Cu. Ternary FePd alloys are largely unexplored in terms of L1 chemical ordering, and results obtained here provide a better understanding of the relation between the identity and quantity of the ternary alloying element and the L1 phase stability and its properties. It is determined that the Curie temperature of the fcc and L1 phases of 3-element-substituted FePd is directly correlated with the average valence concentration, and the trends identified here may be used to predict the effect of other 3-ternary element additions on . Furthermore, it was demonstrated that the structural properties of the L1 phase, as well as the character of the phase transformation in F

ePd modified with small amounts (8 at%)
ePd modified with small amounts (8 at%) of ternary additions, is highly sensitive to the identity of the element incorporated. For instance, while Ni and Cu may be considered very similar elements, additions of Ni do not alter the tetragonality of the L1 phase but quickly reduce the chemical order-disorder temperature. In contrast, Cu additions amplify the tetragonal distortion of the L1 phase but only exert a subtle effect on Likewise, a very narrow L1 stability range at 773 K was established in the Fe-Pd-Ni system, while a wider compositionaprojected in Fe-Pd-Cu. These differences may be related to magnetic interactions in the Fe-Pd-Ni system, which may destabilize the L1 phase. The conclusions of this portion of the dissertation can be extrapolated to the FeNi system, for which the proper ternary alloying selection may favor L1 phase formation in laboratory time scales. In particular, to accomplish this goal elements that increase the of FeNi are required, to either amplify the driving force for nucleation of the L1phase [185] or to permit annealing at higher temperatures to foster increased lattice diffusivities. Since it is unknown which elements may serve to increase of FeNi, a useful starting point would be to select elements that form thermodynamically stable L1Aim 2. Understanding the effects of extrinsic modification, deformation, on the order-disorder A1This Section provides results obtained in f

ulfillment of Aim 2 of this dissertation
ulfillment of Aim 2 of this dissertation. For organization and clarity, this Section has been divided into five parts: Section 4.2.1 provides an introduction and justification for the study; Section 4.2.2 presents specific details of sample synthesis, processing and characterization; Section 4.2.3 provides structural and calorimetric results obtained for this Aim. Specificalpresents results on the effects of cold-rolling on the structural characteristics of the L1FePd phase, and subsections 4.2.3.2 and 4.2.3.3 present kinetic aspects of the chemical ordering transformation in undeformed and deformed samples, respectively; Section 4.2.4 presents an analysis and discussion of the results, while Section 4.2.5 highlights the ficance of this work. In Section 2.1.4.2, previous literature results were presented showing that extrinsic modification delivered to L1-forming ferrous systems through applied mechanical stress or magnetic fields can greatly influence the chemical ordering characteristics as well as impact the magnetic properties of the resultant phase. In particular it was reported that application of stress before or during the chemical ordering transformation accelerates the ordering process, reducing significantly the time required to form the L1 phase at a given temperature. However, the effect of applied external stress on the evolution of the chemical ordering transformation in L1-forming systems needs

greater study to better understand the
greater study to better understand the mechanisms by which plastic deformation enhances the chemical ordering transformation. In this dissertation, studies were performed on the model FePd system to understand the effect of plastic deformation on chemical ordering and disordering. The deformation technique of cold-rolling applied to FePd was chosen as a means to provide a large degree of strain to the alloys, and the effect of this strain on the structural characteristics of the L1 phase, as well as on the kinetics of the L1 ordering transformation, wereFePd sample synthesis, processing and characterization A bulk ingot with a nominal composition Fe was synthesized at Ames high-purity elemental granules that were alloyed and drop-cast into a cylindrical mold with a diameter Ø 1 cm and a length of ~10 cm. A slice of the as-cast ingot, extracted using a diamond saw, was used to verify composition and chemical homogeneity by X-ray fluorescence. The remainder of the drop-cast ingot was then annealed at 773 K for 100 h to induce the A1 transformation. After annealing, a slice of the FePd ingot was extracted using a diamond saw to perform structural characterization through X-ray diffraction, while a 2-mm-thick rectangular section was extracted using electrical discharge machining (EDM, Hansvedt, Ames Lab) for further processing. This rectangular piece was cold-rolled to achieve a total percentage of cold-work of

77%. The effects of plastic deformation
77%. The effects of plastic deformation delivered cold-rolling on the structural characteristics of the L1 phase were investigated by X-ray diffraction performed on both as-annealed and as-cold-rolled samples. Furthermore, analysis of structural and magnetic transformations in these samples upon heating was investigated through constant heating rate differential scanning calorimetry. The effects of plastic deformation through coldtransformation of FePd was also investigated through X-ray diffraction and calorimetry. These studies used two sets of samples: undeformed and deformed FePd. The undeformed FePd samples were extracted from the as-annealed FePd drop-cast ingot and were subjected to a homogenizing treatment at 1050 K for 1 h, followed by quenching at 40 K/min to room temperature. The deformed FePd samples were taken directly from the as-cold-rolled FePd piece. Both sets of samples were characterized by X-ray diffraction to obtain lattice parameters, microstrain and pieces of undeformed and four pieces of deformed FePd samples were annealed isothermally in situ in the DSC to monitor the evolution of the A1 transformation at the isothermal annealing temperatures 490 °C, 495 °C, 500 °C, and 505 °C for a maximum time of 4 h. These temperatures were selected based on the time-temperature-transformation () diagram available for FePd [39] (provided in Section 2.1.2.1), which indicates a maximum rate of che

mical ordering of FePd in the vicinity o
mical ordering of FePd in the vicinity of 525 °C. The DSC signal obtained for the transformation in each sample was analyzed through a JMAK nucleation and growth model (Section 2.2.2.2). Following the isothermal annealing procedure, samples were cooled at a rate of 40 K/min to retain the phase attained though isothermal annealing. The crystal structure of the samples after isothermal treatment was examined by XRD. Results from cold-rolled FePd Effects of cold-rolling on the degThe average composition of the as-cast ingot was verified to be Fe50.449.6 (at%). In the as-annealed state, the sample was confirmed to exhibicorresponding to two phases: an fcc chemically disordered phase and an L1 chemically ordered phase, as demonstrated by the XRD pattern shown in Figure 58a and by the detailed views of selected regions presented in Figure 58b and c. The calculated lattice parameters of the phases are: = 3.807 ± 4 Å for the fcc phase, and = 3.7194 ± 17 Å for the L1 phase, in broad agreement with values reported in the literature for fcc and L1 FePd phases with a composition similar to that of the studied sample [26]. The L1 phase volume fraction, estimated from a relation of the integrated intensities of the Bragg peaks corresponding to each phase (described in Section 3.3.2) is ~86%, with a corresponding 14% volume fraction of the fcc phase. Cold-rolling causes the L1 phase to disorder to the fcc phase, as evi

denced by the X-ray diffraction pattern
denced by the X-ray diffraction pattern obtained for the cold-rolled sample which compares favorably to that reported for equiatomic fcc FePd [175] (Figure 59). Superlattice peaks are not observed within the limits of detection, and the absence of any peak-splitting effects indicate attainment of the cubic symmetry. A processing-induced () preferred orientation texture is observed. The lattice parameter obtained for this cold-rolled sample = 3.820 ± 8 Å, which within experimental error is close to the lattice parameter of the oy prior to cold-rolling. 130 Figure 58. a) X-ray diffractogram obtained for binary FePd annealed at 773 K for 100 h. Reference patterns for fcc FePd (JCPDS 04-003-5130 h. Reference patterns for fcc FePd (JCPDS 04-003-5130 ) and L10 FePd (JCPDS 03-065-9971 065-9971 ) are included for comparison. b) and c) Detailed view of the regions 45° respectively, showing fits corresponding to L1 (blue) and fcc (orange) peaks, together with a cumulative fit (red). Figure 59. X-ray diffractogram obtained for binary FePd annealed at 773 K for 100 h and subsequently cold-rolled. A reference pattern for fcc FePd A reference pattern for fcc FePd ) is included for comparison. Calorimetric investigations of as-annealed and of as-cold-rolled FePd samples indicate an event of second-order character that is associated with the Curie temperature, = 737 K for both samples, Figure 60. Upon heating to high

er temperatures, this event is followed
er temperatures, this event is followed by a broad endothermic event starting at = 770 K, which precedes a sharp endothermic peak. The onset temperatures for the sharp endothermic peak are onset = 946 K and 928 K for the as-annealed and as-cold-rolled sample, respectively. The combination of endothermic events is associated with the complete L1disorder phase transformation, despite the fact that the L1 phase was not detected in XRD for the as-cold-rolled sample. The area under the curve, representing the enthalpy of transformation, is L10 = 23.2 J/g for the as-annealed sample, and is significantly smaller for the as-cold-rolled sample, at L10Figure 60. Differential scanning calorimetry results at 20 K/min on FePd in the as-rolled (orange, bottom) states. To investigate the kinetics of the chemical disorder-order transformation in undeformed FePd, four pieces of the as-annealed FePd alloy were heated to 1050 K, held for 1 h, and subsequently cooled to room temperature at 40 K/min. After this process, the samples were confirmed to consist of a single fcc chemically disordered phase, as demonstrated by X-ray diffraction (Figure 61 shows the diffraction pattern for one of these samples). The lattice parameter was determined as = 3.806 ± 1 Å, consistent with values reported for this composition [26,175]. These samples exhibit negligible microstrain, and were determined to have crystallite sizes � 100 nm.

Figure 61. X-ray diffractogram obtaine
Figure 61. X-ray diffractogram obtained for one of the undeformed FePd samples studied. Reference fcc FePd (JCPDS 04-003-5130 studied. Reference fcc FePd (JCPDS 04-003-5130 ) pattern is Subsequently, the fcc-type samples were subjected to an isothermal treatment in the DSC to promote the fcc transformation, as described in Section 4.2.2. Figure 62 presents the DSC data obtained, where stages preceding the isothermal step (equilibration step at room temperature, constant heating rate up to the isothermal annealing temperature, and time re equilibrate at the isothermal annealing temperature) have been removed. An exothermic DSC peak upon isothermal annealing is observed for all samples, with a peak shape that varies according to the annealing temperature. This exothermic event is attributed to the fccchemical ordering transformation, which takes less time to complete (i. e. for the exothermic signal to return to the baseline) for higher annealing temperatures. Figure 62. Isothermal DSC scans at different temperatures for FePd samples. The DSC signal before the isothermal scan has been removed for The samples have transformed into a single chemically ordered L1 phase after these isothermal annealing treatments, as evidenced by the X-ray diffraction data (Figure 63a). No Bragg peaks corresponding to the fcc phase are observed, demonstrating a complete fcc transformation (see Figure 63b for an enlarged view of

the () split Bragg peaks). This fcc tran
the () split Bragg peaks). This fcc transformation is accompanied by a reduction in the unit cell volume of ~0.1%. The lattice parameters for these samples vary with the annealing conditions, resulting in an decreased ratio with increased isothermal annealing temperature, while their unit cell volumes remain constant within error (Figure 64). b) Figure 63. a) X-ray diffractograms obtained for FePd isothermally annealed in the DSC at 490 °C, 495 °C, 500 °C, 505 °C. Reference L1 FePd (JCPDS [48]) pattern is included for comparison. b) and c) Detailed view of the regions 45° ing peak-splitting indicative of the tetragonal symmetry, and no additional presence of an fcc phase. 135 Figure 64. Variation with isothermal annealing temperature of unit cell volume, and ratio of the L1 FePd phase. The unit cell volume of the starting fcc phase (undeformed, before isothermal treatment) has s are drawn to guide the eye. The evolution of the fcc phase transformation in the FePd system was evaluated from the isothermal DSC data, converted to fraction transformed () as a function of time as described in Section 3.3.4 by integrating the DSC exothermic peak that represents the transformation (Figure 65a). Overall, the fraction transformed into the phase as a function of time shows a sigmoidal character, typical of an isothermal nucleation and growth process. A JMAK kinetic model applied to these data (see Section 2.

2.2.2 for details, Figure 65b) returns A
2.2.2 for details, Figure 65b) returns Avrami exponents and rate constants as presented in Table 6. It can be seen that the Avrami exponent is roughly temperature-independent, with an average value of = 2.33. An Arrhenius plot of ln () vs. 1/ reveals a slope of 34,160 K, from which an effective activation energy associated with the A1 FePd chemical ordering process is determined as ~284 ± 46 kJ/mol (2.94 ± 0.48 eV/atom) Figure 65. a) Isothermal fraction transformed, calculated from the area under the isothermal DSC peak for the transformation, for FePd. b) Linearized data following the JMAK model. ) and intercept (ln ) of the linear fit to the isothermal data representing the chemical ordering transformation in Isothermal temperature 505 °C 2.34 -18.89 0.992 500 °C 2.33 -19.33 0.995 495 °C 2.31 -19.46 0.998 490 °C 2.32 -19.83 0.998 137 Figure 66. Arrhenius plot to determine the activation energy of the chemical Kinetics of chemical ordering: cold-rolled FePd In their as-cold-rolled state, as shown in Section 4.2.3.1, FePd samples exhibit a single fcc phase in XRD data, yet calorimetric detection of an order-disorder L1A1 phase transformation indicates the existence of an L1 phase. These cold-rolled samples exhibit considerable microstrain, 0.26%, and possessed reduced crystallite sizes with an average value of 38 nm. The kinetics of the A1 chemical ordering transformation upon isothermal annealing of t

hese deformed FePd samples at different
hese deformed FePd samples at different the cold-rolled FePd alloy. The heat evolved during the isothermal annealing treatment of these cold-rolled FePd samples is shown in Figure 67, where data corresponding to stages preceding the isothermal step (equilibration step at room temperature, constant heating rate up to isothermal annealing temperature, and time re equilibrate at the isothermal annealing temperature) have been removed. An exothermic DSC signal which decreases in magnitude with increased time is evident as soon as the isothermal step begins. The shape of this exothermic event is very similar for all isothermal temperatures. This calorimetric feature is attributed to the A1 FePd chemical ordering transformation. Figure 67. Isothermal DSC scans for different temperatures for cold-rolled FePd samples. The DSC signal before the isothermal scan has been removed for Isothermal treatment causes a transformation of these cold-rolled samples into the chemically ordered state, as evidenced by X-ray diffraction (Figure 68a). No Bragg peaks corresponding to the fcc phase are observed, demonstrating a nominally complete transformation into the L1 phase (see Figure 68b for an enlarged view of the () split Bragg peaks). Additional low-intensity Bragg refractions may be indexed to a Fe iron oxide. The fcc transformation is accompanied by a reduction in the unit cell volume of ~1%. The lattice parameters for the

se cold-rolled, isothermally annealed sa
se cold-rolled, isothermally annealed samples vary with the isothermal annealing temperature, resulting in an overall increase in ratio with increased annealing temperature, Figure 69; however no significant change in the unit cell volume is noted with increased annealing temperature. Comparison of the lattice parameters obtained from these cold-rolled isothermally annealed L1 FePd samples (Figure 69) to those obtained for undeformed isothermally annealed L1 FePd samples (presented in Section 4.2.3.2, Figure 64) reveals major differences, with higher degrees tetragonal distortions (lower ratiowhen deformation is applied prior to the chemical ordering annealing treatment. Figure 68. a) X-ray diffractograms obtained for cold-rolled FePd isothermally annealed in-situ in the DSC at 490 °C, 495 °C, 500 °C, 505 °C. Reference FePd (JCPDS 03-065-9971 065-9971 ) and Fe3O4 (JCPDS 00-019-0629 (JCPDS 00-019-0629 ) patterns are included for comparison. b) and c) Detailed view of the regions 45° ing peak-splitting indicative of the tetragonal symmetry, and no additional peaks that could signal the presence of an fcc phase. 140 Figure 69. Variation with isothermal annealing temperature of ratio of the L1 FePd phase achieved in cold-rolled samples. The unit cell volume of the starting cold-rolled fcc phase (before isothermal annealing) has been included for reference. Lines are The evolution of the chemical order

ing transformation in deformed FePd was
ing transformation in deformed FePd was investigated from the isothermal DSC data, converted to fraction transformed () as a function of time (Figure 70a). The fraction transformed into the L1 phase first increases rapidly and then plateaus at higher isothermal annealing times. The JMAK kinetic model applied to these data (see Figure 70b) returns Avrami exponents and rate constants as presented in Table 7. The Avrami exponent is roughly constant for all isothermal temperatures, with an average value of = 0.92. An Arrhenius plot of ln ( reveals a slope of 22,765 K that can be used to obtain the effective activation energy associated with the chemical ordering process in cold-rolled samples as ~189 ± 20 kJ/mol (1.96 ± 0.21 eV/atom) (Figure 71). Figure 70. a) Isothermal fraction transformed, calculated from the area under the isothermal DSC peak for the transformation, for cold-rolled FePd. b) Linearized data following the JMAK model. ) and intercept (ln ) of the linear fit to the isothermal data representing the chemical ordering transformation in cold-rolled FePd. The RIsothermal temperature 505 °C 0.91 -6.24 0.997 500 °C 0.92 -6.49 0.998 495 °C 0.92 -6.69 0.995 490 °C 0.91 -6.82 0.996 142 Figure 71. Arrhenius plot to determine the activation energy of the chemical ined for cold-rolled FePd The influence of plastic deformation delivered through cold-rolling on the transformation of the model FePd syst

em is discussed here. Characterization b
em is discussed here. Characterization by structural and calorimetry probes demonstrates that: ) plastic deformation has a strong influence on chemical disordering of the L1 FePd phase, but under the conditions used here does not achieve a fully disordered state, ) plastic deformation reduces the activation energy for the A1 transformation in FePd by 33% in comparison to an unmodified sample, and ) plastic deformation drastically modifies the nucleation and growth mechanisms of the L1 phase in relation to those of the unstrained state. This Section is divided in two parts: Section 4.2.4.1 examines the effects of plastic deformation on the structural character of initially L1 chemically ordered FePd, and Section 4.2.4.2 analyzes the effects of plastic deformation on the chemical ordering phase transformation in FePd samples upon isothermal annealing. Kinetic aspects of the phase transformation are analyzed in the context of nucleation and growth Effects of plastic deformation on the degree of L1The influence of plastic deformation on the degree of chemical order of FePd is discussed in the following paragraphs. A discussion of the structural character of the FePd sample prior to cold-working is presented first, followed by examination of the the degree of chemical order. The FePd annealed sample used for this study was confirmed to consist of a mixture of fcc and L1 phases, with a volume fraction of 14% and 86

%, respectively. This is not the equilib
%, respectively. This is not the equilibrium state, as the Fe-Pd phase diagram (Section 2.1.2.1) displays a phase for the studied composition at the annealing temperature. The results are also in contrast with those presented previously (Aim 1 of this dissertation, Section 4.1.3.1) for a binary FePd of similar composition and annealed under the same temperature/time conditions. Furthermore, as discussed in Section 2.1.2.1, the diagram for the A1 transformation in near-equiatomic FePd predicts a full transformation into the L1 phase at 773 K after a 100 h (annealing conditions). The presence of both fcc and L1 phases in this annealed FePd sample must therefore be related to its processing/characterization history. This FePd alloy was first annealed to achieve the L1 phase, and was sliced afterwards for X-ray analysis. Consistent with the structural observations, it is concluded that cutting the annealed FePd sample induced surface chemical disordering of an otherwise fully L1 ordered alloy, resulting in an phase mixture. The fcc chemically disordered phase is then hypothesized to exist in a thin layer that is concentrated at the cut surface of the sample. However, this phase mixture does not significantly influence the magnetic properties or phase transformation character: the Curie temperature of this annealed FePd alloy is dominated by the majority L1 phase, and the characteristics of the L1A1 phase transfor

mation that occurs in this annealed FePd
mation that occurs in this annealed FePd sample upon heating in the DSC, such as and (Figure 60), are equal within error to those obtained for a sample consisting of Plastic deformation delivered through cold-rolling to the annealed FePd alloy results in the development of a () rolling texture, destruction of long-range chemical order and attainment of a cubic fcc lattice. Nonetheless, regions of L1 chemical order have been indeed retained after cold-rolling, as supported by the observation of an endothermic signal upon heating in the DSC corresponding to a L1A1 phase transformation, Figure 60. The Curie temperature for this cold-rolled FePd sample is identical to that of the annealed FePd alloy, indicating that in this sample is dominated phase. This L1 phase is concluded to be present in nm-sized regions distributed homogeneously throughout the sample, resulting in an X-ray diffractogram dominated by fcc Bragg-peaks, Figure 59. The structural state of the cold-rolled FePd, as described above, is in agreement with previous reports of L1-type FePd subjected to high-pressure torsion (HPT) [33], where the presence of the L1 phase after deformation was asserted only by transmission electron microscopy (TEM) selected area diffraction (SAED) and not by any other structural probe. It is concluded that plastic deformation through cold-rolling induces a strong chemical disordering of the L1 FePd phase, but under the

conditions studied a fully disordered st
conditions studied a fully disordered state is not achieved. The mechanism of strain-induced chemical disordering of FePd is purely mechanical, related to plastic deformation in metals and alloys where dislocation generation and dislocation movement accommodates a permanent deformation. The generation and movement of dislocations in this manner results in atomic rearrangement that destroys the long-range chemical Effects of plastic deformattransformation in FePd Kinetic information pertaining to the nucleation and growth mechanism as well as the energy barrier(s) associated with the A1 phase transformation in undeformed and cold-rolled FePd samples is discussed in the following paragraphs. In the undeformed state, samples are confirmed to consist of a single-fcc phase with negligible microstrain and crystallite sizes above 100 nm, Figure 61. Isothermal annealing at different temperatures induces the A1 transformation, which starts early during the annealing process and lasts for up to ~3.5 h, Figure 62. The transformation goes to completion at all studied temperatures, as confirmed by the presence of a single L1phase after isothermal annealing of undeformed FePd samples, Figure 63. The phase fraction transformed as a function of time during the A1 transformation demonstrates a sigmoidal character, typical of a nucleation and growth process, with a calculated Avrami exponent of 2.33. The growth rate of the L1 p

hase during the A1 phase transformation,
hase during the A1 phase transformation, which is polymorphic in nature (i.e. no compositional change but rather a change in crystal structure), should depend on the rate of atom rearrangement at the interface between transformed and untransformed regions ( interface-controlled growth) [64,126]. In bulk samples there is, in principle, no spatial limit to the growing L1regions, and thus a three-dimensional growth of the L1 phase is expected. Interface-controlled three-dimensional growth of the L1 phase would result in an Avrami exponent between 3 and 4, depending on the nucleation mechanism (Section 2.2.2.2). However, the Avrami exponent obtained for the A1 transformation in the undeformed FePd samples studied here is lower, = 2.33. As described by Barmak et al.[186], lower Avrami exponents can result if, for example, the chemically ordered domains do not remain spherical but rather become elongated as they grow, likely as a result of elastic interactions arising from the change in unit cell volume associated with the transformation. Kinetic studies of the chemical ordering transformation in FePt and CoPt films [125,126,186,187] have also returned smaller Avrami exponents, in the range ~ 1-2. Furthermore, microstructural investigations in the FePd and FePt systems have demonstrated that L1 nuclei emerge during the chemical ordering transformation as thin disks or plates possibly due to the strains associated w

ith coherent nucleation of the tetragona
ith coherent nucleation of the tetragonal L1 phase [23,41,44,45]. The then grow until they impinge on other growing nuclei to produce the polytwinned microstructure discussed in Section 2.1.3.3. The Avrami exponent determined in this dissertation for the A1transformation in undeformed FePd samples is consistent with a site saturated nucleation mechanism of the L1 phase in the A1 matrix, followed by a 2-D interface-controlled The effective activation transformation in bulk undeformed FePd, determined here for the first time as 2.94 ± 0.48 eV/atom, is comparable to the reported activation energy for lattice diffusion in fcc Fein the range 2.41-2.74 eV/atom [188,189]. Similarly, the activation energy for chemical ordering in FePt films and in bulk meteoritic FeNi has been reported to be comparable to that for lattice diffusion in those systems [64,186]. This good agreement between the activation energy for the A1 transformation and that for lattice diffusion implies that the chemical ordering rate depends on the rates of both vacancy creation and vacancy migration, despite the fact that the transformation is expected to proceed via an interface-controlled mechanism. Sufficient energy is required for both the creation and migration of vacancies at the interface (as is true for lattice diffusion) in order to allow the atom rearrangement necessary for the growth of the chemically ordered phase. Examination of the k

inetics of the A1 phase transformation i
inetics of the A1 phase transformation in strained FePd and comparison to results obtained from undeformed samples allows determination of the effect of severe plastic deformation on the evolution of chemical ordering. In the cold-rolled state, as seen in Section 4.2.4.1, FePd samples are confirmed to consist of a majority fcc phase that contains nm-sized L1 regions. Cold-rolling is found to reduce the average crystallite size to values near 38 nm and to induce significant microstrain, on the order of ~0.26%. Isothermal annealing of these cold-rolled FePd samples at different temperatures induces the A1 transformation, which starts as soon as the isothermal DSC segment begins and lasts for ~2.3 h (Figure 67), for an overall reduction in time necessary to achieve the chemical ordering transformation of more than an hour for cold-rolled FePd samples in comparison to undeformed samples. This transformation is confirmed to proceed to completion at all studied temperatures, as revealed by the presence of a single L1 phase after isothermal annealing, Figure 68. The evolution of the fraction transformed into the L1 phase as a function of time during the A1 phase transformation upon isothermal annealing of the cold-rolled FePd samples, analyzed using a JMAK nucleation and growth model, reveals an Avrami exponent of 0.92. Considering interface-controlled growth during the A1 transformation, an overall Avrami exponent o

f ~1 for strained FePd is consistent wit
f ~1 for strained FePd is consistent with one-dimensional growth (see Section 2.2.2.2). In these samples, nm-sized L1 regions were present in the samples before isothermal annealing (see Section 4.2.4.1). These regions are postulated to have served as pre-existing nuclei, which later grew into the L1 phase. Additional nucleation of the L1 phase may have also occurred as a burst of nucleation by virtue of the increased amount of grain boundaries and structural defects produced through severe plastic deformation [190,191]. Such nucleation mechanisms are consistent with an overall Avrami exponent of 1. Notably, the delivery of severe plastic deformation prior to annealing is found to have a definite impact on the growth dimensionality of the L1phase, changing it from two-dimensional to one-dimensional growth. This change in growth dimensionality can be interpreted in the context of the increased number of structural defects present in the deformed samples, at which nucleation and coalescence of nuclei may have taken place, after which a planar front grew to consume the parent phase. In a similar manner, chemical disordering of meteoritic L1 FeNi was determined to occur by nucleation and impingement of disordered A1 grains along interfaces, followed by 1-D growth [64]. Computer simulations and analytical calculations have shown that such mechanism of heterogeneous nucleation followed by a planar-front rall Avrami ex

ponent of 1 [192]. The effective activat
ponent of 1 [192]. The effective activation energy for the A1 transformation in strained FePd, determined here as 1.96 ± 0.21 eV/atom, is much smaller than the activation energy range for lattice diffusion in this system of 2.41-2.74 eV/atom[188,189], and probably closer to the activation energy for grain boundary diffusion in FePd (which was not found in the literature). This determined value implies that less energy is ordering of deformed samples, with a transformation rate that is likely limited by the rate of vacancy migration alone. Vacancy formation may be facilitated in deformed samples due to the increased number of structural defects that results from severe plastic deformation [190,191]. The lower activation energy for phase transformation found in deformed samples indicates an enhancement in the rate of chemical ordering, in comparison with that of undeformed samples, consistent with reports of reduced times required to attain the L1 phase in deformed FePd in comparison with conventionally-ordered FePd [35,113,115,116,193]. Finally, the structural characteristics of the L1 phase formed upon isothermal annealing of undeformed and strained samples are notably different, Figure 64 and Figure 69. Deformation induces the formation of an L1 phase with much higher lattice parameter and much smaller lattice parameter than those obtained in undeformed samples, resulting in an overall lower ratio. This di

fference must certainly be related to th
fference must certainly be related to the significant amount of strain present in the lattice of the deformed samples, which affects the final characteristics of the L1 phase, and likely impacts the intrinsic magnetic properties of the deformed FePd alloSignificance of work on processing-structure correlations in FePd The conclusions on the effect of cold-working on the A1 phase transformation in FePd are summarized as follows. Cold-working of L1 FePd results in the emergence of a crystallographic texture, reduction in crystallite size and increase in microstrain, accompanied by significant destruction of long-range order. Nonetheless, complete chemical disordering is not attained through cold-workistudied in this dissertation. The mechanism by which plastic deformation destroys the long-range order is purely mechanical, related to dislocation generation and movement which quickly, but not completely, transforms the L1 lattice into a chemically disordered The growth of the L1 phase in cold-worked FePd proceeds in a one-dimensional manner, as opposed to the growth mechanism operating in undeformed samples, which was determined to be two-dimensional. This difference in growth mechanism is attributed to the increased number of structural defects existing in deformed samples, which facilitate heterogeneous nucleation of the L1 phase followed by coalescence of those ergies for chemical ordering in undeformed and str

ained FePd are very different, calculate
ained FePd are very different, calculated for the first time here as 2.94 ± 0.48 eV/atom and 1.96 ± 0.21 eV/atom, respectively. The activation energy for the A1transformation in undeformed FePd is consistent with a rate of phase transformation that is limited by the requirement to create and move vacancies at the interface. The much lower activation energy for the A1 transformation in cold-rolled FePd suggests that the highly defective microstructure generated by severe plastic deformation facilitates vacancy formation at the interfaces, and the rate of the transformation is then limited by the requirement to move the vacancies to achieve rearrangement of atoms into a chemically ordered arrangement. These results, in agreement with qualitative reports of enhanced kinetics for severely deformed FePd in comparison with undeformed FePd [35,113,115,116,193], can be extrapolated to the related FeNi systems, for which processing routes to accelerate the chemical ordering kinetics are actively under investigation. Application of severe plastic deformation to bulk FeNi is anticipated to facilitate the A1 chemical ordering process, and constitutes a suitable technique for the laboratory synthesis of the highly interesting L1 FeNi phase in industrially-relevant timescales. This Section provides results obtained in fulfillment of Aim 3 of this dissertation. For organization and clarity, this Section has been divided int

o five parts: Section 4.3.1 provides an
o five parts: Section 4.3.1 provides an introduction and justification for the study; Section 4.3.2 presents specific details of sample synthesis, processing and characterization; information pertaining samples used for synchrotron X-ray diffraction characterization is presented in subsection 4.3.2.1 , and that of samples used for neutron diffraction characterization is presented in subsection 4.3.2.2 ; Section 4.3.3 , also subdivided in two subsections, provides structural and magnetic results on the materials studied; Section 4.3.4 presents analyses and discussion of the results, and thereby provides an increased understanding of the transformation in the FeNi system. Finally, Section 4.3.5 highlights the As described in Sections 2.1.4, previous literature reports that substitutional ternary alloying additions and severe plastic deformation applied to ferrous L1-forming systems can be used to tailor the chemical ordering transformation and the magnetic properties of the resultant L1 phase. Furthermore, Sections 4.1 and 4.2 of this dissertation demonstrated that L1 phase formation in FePd-based ternary alloys is highly sensitive to the identity of the ternary element added, and that cold-rolling processing applied to the model FePd system prior to the chemical ordering annealing treatment has a profound effect on the mechanism of nucleation and growth of the L1 phase and reduces the activation energy for the A

1 transformation. In the FeNi system, ev
1 transformation. In the FeNi system, even though it has been reported that the rate of chemical ordering is vacancy-controlled [64,68], efforts to produce excess vacancies through severe plastic deformation for the promotion of the L1 phase have produced yet inconclusive results, as was described in An estimation of the excess vacancy concentration in the FeNi system required to facilitate the achievement of the L1 phase can be made comparison with another system that undergoes L1 chemical ordering at a similar order-disorder temperature. The Gold-Copper (AuCu) system is the closest option, with an order-disorder temperature of 658 K [194]. In the AuCu sytem, which has a lattice diffusivity of ~1x10-14 /s at 573 K [195], typical annealing times in the range of 8 to 14 days are used to produce the phase in the temperature range 573-616 K [196]. In FeNi, the lattice diffusivity of the equiatomic composition at 573 K is reported as ~1x10-27 /s [69], and the L1 phase forms over periods of millions of years. It is thus anticipated that for L1 phase formation to be possible in FeNi in laboratory time scales, the lattice diffusivity of this system must be increased by a factor of 10. Since diffusivity is directly related to the vacancy concentration, for chemical ordering to proceed in FeNi similarly as it does in AuCu the concentration of vacancies must also be increased by ~10, provided the activation energy for

atom migration is kept constant. formati
atom migration is kept constant. formation in FeNi as that reported for the formation of a vacancy adjacent to Fe in a Ni host (1.7 eV) [185], the equilibrium vacancy concentration in FeNi at 573 K can be estimated as ~1x10-15/mol. It has been shown that the application of plastic deformation techniques can produce vacancy concentrations in metals and alloys that are close to the equilibrium vacancy values at the melting temperature [153,197–199], which for FeNi is estimated here as ~1x10/mol. If this is the case, plastic deformation applied to FeNi has the potential to increase the vacancy concentration by ~10vacancies per mol. This analysis indicates that the concentration of vacancies produced by plastic deformation is insufficient to achieve chemical ordering to form L1-type FeNi in a fean increase in the vacancy concentration by a factor of 10as a result of plastic deformation may significantly reduce the chemical ordering time in FeNi from millions of years to a laboratory time scale. Of course, this would be the case provided that the introduced vacancies do not annihilate via migration to sinks such as dislocations (also produced extensively by plastic deformation, to the order of [199,200]), and that therefore they can effectively be used to facilitate chemical ordering. The slocations, which can be roughly estimated as 1/m, indicates that in a severely deformed metal or alloy vacancies need to travel

a distance of ~5-50 nm to reach a dislo
a distance of ~5-50 nm to reach a dislocation and annihilate. In the FeNi system, considering the sluggish diffusion at low temperature, it is possible that vacancies produced by plastic deformation at cryogenic temperatures or room temperature can be retained to favor a subsequent chemTo contribute useful information for development of the understanding of the effects of chemical and microstructural modifications on the structure and magnetism of FeNi alloys, as well as on the characteristics of the A1 phase transformation, FeNiM (M = Ti, Al, V) alloys subjected to severe plastic deformation were studied in this dissertation. Ternary alloying additions of Al, V, and Ti were selected on the basis of their negative heats of formation for an L1 phase with Fe or Ni [147], and severe plastic deformation by cold-rolling and cryomilling was used to deliver large amounts of strain and increased number of extrinsic structural defects to the FeNi-based alloys, to enhance atomic diffusion at low temperatures. FeNi-based sample synthesis, processing and characterization Two different sets of samples were fabricated for the experiments pertaining to Aim 3. The first set corresponds to samples that were probed usiray diffraction. The second set corresponds to samples that were probed using neutron diffraction. A description of synthesis, processing, and characterization conditions for both sets of samples is provided sepa

ratediffraction experiments master allo
ratediffraction experiments master alloy was used as the starting material for all alloy compositions. Ternary alloying additions of M (M = Al (99.999%), Ti (99.99%), or V (99.8%)) were incorporated into the master alloy to produce arc-melted (FeNi)ingots. Afterwards, Fe and (FeNi) ingots were melt-spun into ribbons with typical dimensions 15 cm long, 3 mm wide, and 200 µm thick. The nominal composition and chemical homogeneity of the as-melt-spun ribbons were verified using SEM/EDS. The ribbons were cut to 5 mm lengths, and the resulting pieces (~1 g for each composition) were cryomilled inside a liquid nitrogen bath (Section 3.2.2). To minimize oxidation of the FeNi-based samples during the cryomilling process, a surfactant mixture of oleic acid (25 wt%) mixed in heptane (25 wt%) was used to provide a protective coating layer to the materials, and the vials were sealed inside a glovebox with an Ar atmosphere. Processing cycles were composed of 10 min of milling at a rate of 15 cycles per second followed by 2 min of cooling. After optimized cumulative milling times of 9 h (see Appendix 4 for more information), the samples (in powder form) were collected and rinsed with acetone and heptane in order to remove the surfactant. Samples were then encapsulated in silica tubes evacuated to a nominal phase formation. Structural examination of the FeNi-based samples in the as-milled and as-annealed states was conduct

ed using synction at the National Synchr
ed using synction at the National Synchrotron Light Source NSLS (BNL) Beamline X-16C. The powder samples were probed in rotating silica capillaries with a selection of incident X-ray energies/waveleghts. An initial experiment, carried out at a wavelength of = 0.700292 Å, focused on detecting any possible peak-splitting that could indicate tetragonality in the crystal structure. A point detector incorporating a Ge()-analyzer was used to obtain a high angular resolution. A follow-up experiment, performed on one sample of interest, focused on detecting the () superlattice peak indicative of chemical ordering. This peak was selected considering that it is the most intense superlattice peak in untextured L1 FeNi, and that it does not coincide with Bragg reflections from easily-formed oxides. The experiment was performed under anomalous diffraction conditions (see Appendix 2) to amplify the contrast between Fe and Ni to achieve an enhanced intensity of the superstructure Bragg peaks relative to that of the fundamental reflections. These anomalous diffraction experiments were performed with incident beam wavelengths in the range 1.74700 Å 1.87220 Å, near the Fe-K absorption edge. In this configuration, a high-sensitivity Si-strip detector was used. Additional samples were synthesized to further probe the effect of severe plastic deformation and subsequent annealing on FeNi-based alloys, and to assess the potential o

f these techniques for achievement of L1
f these techniques for achievement of L1 FeNi. Samples of nominal composition and (FeNi) were fabricated by arc-melting or drop-casting natural Fe, Ni, and Ti metal sources, or natural Fe and Ti together with isotopic Ni (performed at Ames Ni was supplied by the United States Department of Energy Office of Science by the Isotope Program in the Office of Nuclear Physics. The Ni was originally in powder form, and prior to arc-melting it was sintered at 1073 K in an atmosphere of Ar-5% H for 2 hours. All synthesized samples were thermally homogenized at 773 K for 100 h prior toA melt-spinning and subsequent cryomilling process, as described in Section 4.3.2.1, was applied to an alloy of nominal composition (FeNi)(referred to as FeNi(Ti) cryomilled). Additionally, all synthesized alloys were cold-rolled to deliver percentages of cold work in the range 83% - 94%. These samples are referred to as FeNi cold-rolled, FeNi(Ti) cold-rolled, FeNi cold-rolled, and FeNi(Ti) cold-rolled. Deformed samples (cryomilled and cold-rolled) were encapsulated in evacuated (10Torr) silica tubes and annealed at 563 ± 5 K for 24 days (samples containing natural Ni) or 30 days (samples containing isotopic Ni) to induce phase transformation. The composition and chemical homogeneity of the samples were verified using energy-dispersive X-ray spectroscopy (SEM-EDS) or Preliminary structural characterization of all alloys in their as-made, as-

deformed, and as-annealed states was ac
deformed, and as-annealed states was achieved by laboratory X-ray diffraction. High resolution time-of-flight neutron powder diffraction measurements were conducted for selected samples at the ISIS pulsed neutron source facility, beamline HRPD (Rutherford Appleton Laboratory, UK). Samples were placed in 6 mm cylindrical vanadium cans for room temperature scans. Data from processed samples made from natural metal sources were acquired with neutron wavelengths in the range 1.24 – 8.24 Å. Data from unprocessed specimens made from natural Ni, as well as processed FeNi-based samples, were collected with neutron wavelengths in the range 1.24 – 5.36 Å. Data were then exported and refined using the suite of software GSAS/EXPGUI [201,202] in collaboration with Dr. Dominic Fortes (Rutherford Appleton Laboratory) and Dr. Luke G. Marshall (Northeastern University). Magnetic characterization of a square FeNi(Ti) cold-rolled sample was performed using a MPMS SQUID magnetometer before and after annealing. In-plane were collected at = 10 K with a maximum applied field of 40 kOe; measurements were done both parallel and perpendicular to the rolling direction. The correction factor for self-demagnetization effects was obtained from the measurement of a Ni standard with the same geometry as the FeNi(Ti) samples. FeNi and FeNiM (M = FeNi and FeNiM (M = Al, Ti, V) struThe compositions of the as-melt-spun ribbons in an atomic per

cent basis were verified to be Fe50.949.
cent basis were verified to be Fe50.949.148.348.53.248.31.5 and Fe49.049.61.4. These samples will be henceforth referred to as FeNi, FeNi(Al), FeNi(V) and FeNi(Ti), respectively. Within instrumental resolution, all compositions in the as-milled and as-annealed states exhibit Bragg reflections consistent with the chemically disordered fcc in the limits of detection (Figure 72). Figure 72. Synchrotron X-ray diffractograms for all compositions in their as-milled Incorporation of transition-metal alloying additions into the parent FeNi alloy provides a small but finite unit cell volume variation that may be best visualized in Figure 73, as a function of Fe content in at%. These results confirm that are indeed incorporated into the FeNi lattice. The results obtained for the binary FeNi sample are consistent with those reported in literature for an fcc phase of this composition [177]. Additionally, Figure 73 displays unit cell volumes reported for meteorite-derived tetrataenite (L1 FeNi) [62]; it can be seen that the unit cell volume of all samples of this study is approximately 0.7% greater than that of meteorite-derived L1 FeNi. Further, the unit cell volumes are expanded upon annealing for the FeNi and FeNi(V) alloys, but experience no change for other compositions. Figure 73. Variation of unit cell volume as a function of Fe content (at%). Reference values for fcc FeNi are included and connected by a linear

fit (dashed line) ected by a linear fit
fit (dashed line) ected by a linear fit (dashed line) . Reference values for meteorite-derived L1 FeNi are also included . Data derived from the Williamson-Hall analysis reveal that the crystallite sizes, in the range 54-80 nm, remain unchanged between the as-milled and annealed states within experimental error. The lattice strain present in the as-milled samples is substantial (0.28% - 0.35%) and is relieved to values in the range 0.22% - 0.29% upon annealing. No llite size or strain with composition. Examination of the synchrotron X-ray diffraction data indicates no peak splitting for the (), () and () Bragg peaks in all as-milled and as-annealed samples. However, when going to larger values and examining the () Bragg reflection, an apparent peak splitting is detected for the FeNi(Ti) annealed sample, as shown in Figure 74. For this sample, two br = 45.9° and 46.1°, and are consistent with the () peak splitting expected for L1 FeNi. These results are in agreement with the higher -spacing resolution achievable at higher detection of peak splitting for the higher-symmetry for this particular sample. Utilizing the position of these split-peaks, together with those of all other Bragg reflections identified in this sample, a least-square fitting procedure to a tetragonal lattice results in lattice parameters = 3.5826(23) Å and = 3.5914(26) Å. This indicates a tetragonal axial ratio = 1.0025(10), which is com

parable to values reported for meteoriti
parable to values reported for meteoritic L1ffractograms obtained with = 0.700292 Å radiation for a) milled samples, and b) annealed samples, showing the region near the () Bragg peak. An X-ray diffraction pattern for L1FeNi, calculated from lattice parameters reported by Albertsen for reported by Albertsen for , is included for comparison. Further investigation of the presence of L1 chemical ordering was performed in the FeNi(Ti) annealed sample alone, using the technique of anomalous diffraction. Different wavelengths near the Fe-K absorption edge were selected and the region for superlattice peak is anticipated was probed. In addition, the (fundamental Bragg reflection was also examined. As presented in Figure 75a, it was not possible to detect a ( superlattice peak in this sample under the experimental conditions used. A signal is observed in the region 30° 2 32° which is attributed to an instrumental artifact, considering that there is no systematic shift in position with systematic variation in X-ray wavelength as would be expected for a Bragg peak. The ) fundamental peak (Figure 75b) was observed to shift in position with varying s obtained under anomalous diffraction conditions for the (FeNi)Ti annealed sample, with wavelength in the range region where superlattice (would be expected, and b) peak. FeNi and FeNi(Ti) structural resumagnetic results To further probe the tetragonality induced thro

ugh severe plastic deformation in the Fe
ugh severe plastic deformation in the FeNi(Ti) annealed sample detected (Section 4.3.3.1), additional samples were synthesized and processed by cryomilling or cold-rolling, (described in Section 4.3.2.2) for neutron diffraction studies. Some of these samples were fabricated with isotopic Ni to enhance any nuclear scattering signal originating from long-range chemical ordering. The assessed compositions in (at%) were (referred to as FeNi, cold rolled), Fe (referred to as FeNi(Ti), cold rolled), Fe (referred to as FeNi(Ti), cryomilled), Fe (referred to as Ni, cold rolled) and Fe (referred to as FeNi(Ti) cold rolled). Samples were investigated initially by laboratory X-ray diffraction and subsequently by neutron diffraction at the ISIS pulsed neutron source. Initial structural characterization of precursor as-made alloys (i.e. melt-spun, drop-cast or arc-melted) and of processed ( deformed pre-anneal and deformed post-anneal) samples conducted with laboratory X-ray diffraction showed, within equipment resolution, that all samples crystallize in the fcc structure. Lattice -parameters were found to be in the range 3.586 Å – 3.627 Å, in broad agreement with the reported range of -parameters for FeNi alloys [177], and reflective of the slight compositional variations. Determined dimensions of the coherently scattering domains, or crystallite sizes, after deformation processing but prior to annealing were typically u

nder 100 nm. Upon annealing, the crystal
nder 100 nm. Upon annealing, the crystallite sizes remained relatively constant for all samples, irrespective of the type of deformation processing. Deformation processing produced significant Brabroadening, associated with microstrain levels in the range 0.18% - 0.90%. The degree of initial strain present in the samples after severe plastic deformation processing is highly dependent upon the particular processing method employed. Further, X-ray diffraction data obtained from cold-rolled unannealed FeNi and FeNi(Ti) samples contained coexisting “doubled” Bragg peaks – each sharp Bragg peak is partnered with a broad Bragg peak – consistent with the existence of two phases of differing lattice parameters, crystallite sizes and microstrain levels. The multiphase nature of these cold-rolled samples disappeared after annealing. These data are available in Table 8. Table 8. Microstructural conditions of FeNi samples determined from laboratory X- = cubic lattice parameter, = coherently-diffracting crystallite size, = microstrain. condition Undeformed Deformed, pre-anneal Deformed, Cryomilled = 3.5870(5) Å � 100 nm =0.014 ± 0.006% = 3.586(1) Å = 64 ± 14 nm = 0.30 ± 0.03% = 3.586(1) Å = 60 ± 15 nm = 0.23 ± 0.04% = 3.599(4) Å = 50 ± 16 nm = 0.19 ± 0.07 % Bulk values Surface values = 3.600(3) Å = 36 ± 12 nm = 0.17 ± 0.10% = 3.596(3) Å = 51 ± 18 nm = 0.21 ± 0.07% = 3.627(3) Å = 16 ± 8 nm

= 0.82 ± 0.47% = 3.592(2) Å � =
= 0.82 ± 0.47% = 3.592(2) Å � = 100 nm =0.016 ± 0.04 % = 3.593(2) Å = 63 ± 20 nm = 0.23 ± 0.05 % = 3.595(2) Å = 61 ± 21 nm = 0.20 ± 0.06 %46 = 3.592(4) Å = 40 ± 13 nm = 0.08 ± 0.09 % Bulk values Surface values = 3.592(3) Å = 80 ± 34 nm = 0.22 ± 0.07% = 3.588(2) Å � 100 nm = 0.21 ± 0.04 % = 3.627(9) Å = 24 ± 14 nm = 0.89 ± 0.36% = 3.586(7) Å � 100 nm = 0.004 ± 0.03 % = 3.588(1) Å = 60 ± 24 nm = 0.18 ± 0.08 % = 3.590(2) Å = 72 ± 27 nm = 0.14 ± 0.06 % High-resolution time-of-flight neutron powder diffraction data obtained from samples made from natural metal sources are discussed first, followed by discussion of data obtained from samples fabricated from isotopic Ni. Samples made from natural metal sources were characterized in their as-made (undeformed) and as-processed (deformed post-anneal) states. Data for the deformed pre-anneal state is not available. erns obtained for all examined samples show Bragg reflections that, to the naked eye, correspond to an fcc phase. A minor secondary phase of FeO was detected, at an abundance 1 wt%. Rietveld refinements of these data employed both fcc and face-centered-tetragonal (fct) structural models (see Figure 76). While refinements for unprocessed samples possessed similar goodness-of-fit parameters for both structural models, the fct model returned a maximum calculac/a ratio, of 1.00076(1). This represents a maximum t

etragonal distortion of 0.08%. By compar
etragonal distortion of 0.08%. By comparison, refinement results from application of the fct model to data obtained from processed samples, provide a minimum calculated tetragonality of = 1.00241(6), representing a tetragonal distortion of approximately 0.24% (Table 9). For better visualization of the data, the degrees of tetragonality (ratios) determined with the fct models for all samples are shown in Figure 77 as a function of Fe content. Values from naturally-occurring tetrataenite, reported by Albertsen [62] for four different meteorites (Odessa IA, Toluca IA, Dayton IIID, Mincy Mesosiderite) and confirmed for the NWA6259 meteorite [40], are included for reference. Also included are the results of the refinements for a control sample of a commercially available Fe powder (Alfa Aesar, -325 mesh, 99.5+% purity). Figure 77 shows that the axial ratios of all unprocessed samples fall below a value of ~ 1.0008 while those of the processed samples fall above a value of Further, the values determined from data obtained from processed samples are consistent with those of meteoritic L1 FeNi, while the values determined from unprocessed samples are consistent with that of commercially available fcc FeNi powder. The degrees of tetragonality of the processed materials and of the meteoritic-derived tetrataenite exhibit little variation around the equiatomic composition and then decrease with increased iron content to

a value of 1.0020. In contrast, the degr
a value of 1.0020. In contrast, the degree of tetragonality of the unprocessed samples shows little 166 cold-rolled annealed FeNi(Ti). Observed data points are shown with (bottom) and without (top) the calculated Rietveld fits overlaid. Differences between the observed and calculated patterns are displayed in the cyan trace located below the peaks. Detailed views of the b) () fundamental Bragg peak, and c) () fundamental Bragg peak, derived from unprocessed FeNi(Ti) (red trace) and cold-rolled, annealed FeNi(Ti) (black trace), illustrating processing-induced Table 9. Rietveld refinement results for commercially available FeNi powder (Alfa Aesar), unprocessed FeNi and FeNi(Ti) samples, and processed FeNi and FeNi(Ti) samples. The goodness-of-fit and lattice parameters for each sample are shown for both the face-centered cubic (fcc) and face-centered tetragonal (fct) models. is the goodness of fit, R is the weighted-profile is the residual of least-squares refinement, and are lattice parameters, is unit cell volume, and is a metric of the tetragonality. Sample type 2 Rwp (%) (%) (Å) (Å) V (ÅAlfa Aesar Powder fcc 7.625 6.25 5.07 3.587135(5) 3.587135(5) 46.1576(1) 1 fct 7.868 6.35 5.43 3.58658(4) 3.58753(6) 46.148(1) 1.00027(2) Drop-Cast fcc 3.915 9.19 8.45 3.586581(8) 3.586581(8) 46.1362(2) 1 fct 3.127 8.22 7.37 3.58593(4) 3.58722(7) 46.128(2) 1.00036(2) Drop-Cast fcc 5.332 11.08 9.71 3.59198(2) 3.59

198(2) 46.3449(3) 1 fct 4.898 10.62 9.7
198(2) 46.3449(3) 1 fct 4.898 10.62 9.75 3.59170(3) 3.59200(6) 46.3380(9) 1.00008(2) As-Spun fcc 1.421 9.23 7.35 3.588754(6) 3.588754(6) 46.2201(1) 1 Ribbons fct 1.432 9.27 8.01 3.58758(3) 3.59029(4) 46.2099(7) 1.00076(1) Cold-rolled fcc 1.660 8.48% 9.04% 3.58655(3) 46.1351(5) 1 1.126 6.98% 7.09% 3.5835(1) 3.5921(2) 46.128(3) 1.00241(6) Cold-rolled fcc 1.588 8.11% 7.59% 3.59235(3) 46.3592(6) 1 fct 1.445 7.73% 7.56% 3.5886(2) 3.6002(3) 46.363(4) 1.00325(7) Cryomilled fcc 1.208 6.41% 6.48% 3.58850(3) 46.2103(6) 1 fct 1.146 6.25% 6.30% 3.5843(1) 3.5968(2) 46.209(3) 1.00348(6) Figure 77. Sample tetragonality as a function of Fe content, obtained from Rietveld refinements that employed an fct structural model to the neutron diffraction data of samples made from natural metal sources. Published data from natural tetrataenite are included data from natural tetrataenite are included . The shaded region indicates a band of values that sepasamples. The variation in the unit cell volumes as a function of Fe content is displayed in Figure 78, where it can be seen that the unit cell volumes of all synthesized samples are approximately 0.7% greater than those of all meteorite-derived tetrataenite (L1 FeNi phase). Further, there is very little difference between the unit cell volumes of the unprocessed samples and those of the processed samples. The Ti-containing samples consistently exhibit an expanded unit cell l

attice relative to the Ti-free samples.
attice relative to the Ti-free samples. It is also observed that the volume obtained for the binary FeNi sample is slightly off from the trends reported for fcc FeNi, as is the volume of the commercially available FeNi Figure 78. Unit cell volume as a function of Fe content, obtained from Rietveld refinements that employed an fct structural model to the neutron diffraction data of samples made from natural metal sources. Published data from natural tetrataenite are included e included , as well as for fcc FeNi . High-resolution time-of-flight neutron powder diffraction was also performed on the isotopically-modified samples in their deformed pre-annealed and deformed post-annealed states. Data for the as-made (undeformed) state is not available. Rietveld refinements of these data employed only thmodels, Table 10. The tetragonal ratios determined for the deformed post-anneal samples FeNi and FeNi(Ti) were determined as 1.0026(3) and 1.0028(3), respectively. c/a ratio values are identical (within error) to those of the post-processed samples made from naturally occurring FeNi. The order-of-magnitude larger uncertainties in these values are attributed to the greatly decreased fundamental Bragg peak scattered intensities from the isotopically-modified samples that arises from the near null-scattering combination of Fe and Ni neutron scattering lengths in the absence of long-range order. The maximum tetragonality

of samples in the deformed pre-anneal s
of samples in the deformed pre-anneal state is = 1.00055(19). This value confirms that the onset of tetragonality does not occur until Table 10. Rietveld refinement results employing the tetragonal fct model for deformed pre-anneal and deformed post-anneal FeNi and Fe is the goodness of fit, R is the weighted-profile R-factor, Rthe residual of least-squares refinement, and are lattice parameters, is unit cell volume, and Sample type 2 Rwp (%) (%) (Å) (Å) V (ÅCold-rolled pre-0.1444 3.46 2.38 3.5850(3) 3.5870(6) 46.102(9) 1.00055(19) Cold-rolled post-fct 0.1792 3.76 2.70 3.5826(6) 3.5920(10) 46.104(16) 1.00262(32) Cold-rolled pre-fct 0.1898 3.93 2.69 3.5905(3) 3.5916(3) 46.303(6) 1.00029(11) Cold-rolled post-fct 0.1646 3.73 2.58 3.5879(6) 3.5980(10) 46.316(17) 1.00283(32) While tetragonality is confirmed in the processed samples, it is determined that superstructure peaks are absent from neutron diffraction data of samples made of Ni. Therefore, the detected phase is a new tetragonal chemically disordered phase (A6 crystal structure) in the FeNi system However, a diffuse scattering signal was clearly identified in all isotopically modified samples, which consisted of a low-intensity, semi-regular waveform concentrated at low -spacings (Figure 79). This diffuse scattering signal was most clearly observable in the deformed post-anneal FeNi sample, and was confirmed to be absent in the initial set

of experiments on samples made from nat
of experiments on samples made from natural metal sources. tterns of a) cold-rolled FeNi sample, and b) cold-rolled annealed FeNi sample. A semi-regular waveform background that corresponds to a diffuse scattering signal is observed in both patterns. To evaluate changes in the character of the magnetic response of the FeNi-based samples upon processing, the magnetization of the FeNi(Ti) cold-rolled sample was measured at = 10 K to assess the approach to magnetic saturation manifest before and after conductance of the annealing protocol. As shown in Figure 80, an appreciable annealing-induced decrease in the initial susceptibility (slope of ) curve) is confirmed for data collected from deformed and annealed samples, measured in orientations both parallel and perpendicular to the rolling direction. Further, a 6% decrease in the saturation magnetization is noted in this sample after annealing. 171 Figure 80. Magnetization data obtained at 10 K for the cold-rolled FeNi(Ti) sample measured before and after annealing. Data were collected in the in-plane cular to the rolling direction. r FeNi and FeNiM (M = Al, Ti, V) In this Section, results pertaining the effects of plastic deformation and ternary alloying additions on the A1 transformation in FeNi are discussed. It is demonstrated that a tetragonal, chemically disordered phase in near-equiatomic FeNi-based alloys is consistently achieved through plastic defo

rmation and annealing. This phase has no
rmation and annealing. This phase has not been reported to date for the FeNi material system. Analyssevere plastic deformation is presented first (Section 4.3.4.1), including a discussion of the mechanism proposed for the A1 transformation of FeNi. Next, the effects of of FeNi are examined (Section 4.3.4.2). Effects of severe plastic deformation on the A1The influence of plastic deformation on the A1 transformation of near-equiatomic FeNi-based alloys is examined in the following paragraphs. Examination of the structural state of all samples in their undeformed, deformed pre-anneal and deformed post-anneal states is presented first, followed by a discussion of an unreported A6 phase obtained consistently in the studied samples through plastic deformation and annealing. The magnetic signature of this A6 phase is presented next, and the Section closes with a proposed mechanism of the A1 transformation in FeNi. It was demonstrated in Sections 4.3.3.1 and 4.3.3.2 that structural features associated with a tetragonal chemically disordered phase are present in FeNi-based samples exposed to deformation and annealing. Synchrotron X-ray a FeNi(Ti) cryomilled and annealed sample (Figure 72 and Figure 74) revealed the presence of one Bragg reflection that does not correspond to an fcc lattice, and that is consistent with peak-splitting originating from a tetragonal symmetry, which can be detected more easily at higher s

cattering angles. There was no signature
cattering angles. There was no signature of peak-splitting in this sample prior to annealing. Neutron diffraction results for all tested FeNi and FeNi(Ti) compositions revealed onal structural model can be used to describe the alloys in their undeformed, deformed pre-anneal, and deformed post-anneal states, with similar goodness-of-fit parameters obtained for both models (Table 9 and Table 10). Nonetheless, the maximum percentage of tetragonawhen using a tetragonal model to describe samples in their undeformed or deformed pre-anneal states, was more than three times smaller than that obtained for the deformed post-anneal samples, Figure 77. The lower limit on difference in ratio between the processed and unprocessed samples is 0.00165(6), which is approximately 9 times larger than the uncertainty on the difference. Thus, while small, this difference is highly statistically significant. Therefore, it is concluded from the structural analysis that a cubic structure is most appropriate to describe near-equiatomic FeNi and FeNi(Ti) samples in the undeformed and deformed pre-anneal states, while the tetragonal structure is confirmed as most appropriate for the samples in the deformed post-anneal state. The absence of superstructure peaks in for all studied samples, including alloys synthesized with Ni (Figure 75 and Figure 79), unequivocally confirms the absence of long-range chemical order in both the cubic (und

eformed, deformed pre-anneal) and tetrag
eformed, deformed pre-anneal) and tetragonal states (deformed post-anneal). However, the presence of a diffuse scattering signal observed in all isotopically-modified alloys suggests the presence of local structural correlations. Diffuse scattering was not detected in natural-Ni containing samples, and thus a relation can be drawn between the occurrence of this signal and the use of the isotopic Ni to synthesize the alloy. Therefore, these proposed local correlations must be closely related to chemical short-range order phenomena. In fact, diffuse scattering signals obtained in neutron scattering experiments carried out on isotopically-modified Fe samples were used to examine short-range order in permalloys [203]. In agreement with analogous observations by et al. from electron diffraction patterns of equiatomic FePd single crystals subjected to annealing and quenching processes [45,46], this diffuse scattering signal may be associated with the formation of a gonal state with no long-range order but a certain degree of short-range order. Detailed analysis of the diffuse requires a different experimental set-up. Overall, the detected tetragonality in near-equiatomic FeNi-based bulk samples, together with the confirmed absence of long-range chemical ordering, establishes that severe plastic deformation followed by proper annealing protocols promotes the synthesis of a newly discovered phase with the tetragonal ch

emically-disordered (A6, space group ) s
emically-disordered (A6, space group ) structure in the FeNi system. As discussed by Vlasova et al. [46], the A6 structure may exist as a symmetry-permitted transitional phase that links the disordered cubic A1 phase with the chemically ordered tetragonal L1 phase in the closely related near-equiatomic FePd alloy, and likely in other L1 forming systems. In fact, Orehotsky [204–206] report the existence of a transient state between the A1 and L1 phases in bulk CoPt samples. Confirmed by X-ray diffraction studies as a non-uniformly strained fcc state [204], it is possible that this identified CoPt transient phase could in fact possess the tetragonal A6-type symmetry. Furthermore, a transitional magnetic state has been observed in meteoritic L1 FeNi upon cyclic heating/cooling to induce the L1transformation [207], Figure 81. This intermediate magnetic state exhibits intrinsic an extrinsic magnetic properties in-between those of L1 and A1 FeNi phases, and thus it is 175 Figure 81. Magnetization of FeNi obtained from the NWA6259 meteorite as a function of a) applied field, and b) temperature. In the as-received state, the meteorite is confirmed to be in the L1 state, while after two heating/cooling cycles between room temperature and 700 °C, the sample is confirmed to be in the A1 state. There is an intermediate magnetic state attained after the first heating/cot heating/co Plastic deformation imparted by cryomill

ing and cold-rolling, which produced sim
ing and cold-rolling, which produced similar microstrain levels in the near-equiatomic FeNi and FeNi(Ti) samples studied (Table 8), is confirmed to have the same effect on the structural character of the A6 phase in these processed samples, achieving comparable degrees of tetragonality. The ratio attained for the A6 phase in deformed post-anneal FeNi-based samples is in close agreement with the tetragonality reported for meteoritic L1 FeNi (average [40,62], while the unit cell volume of the A6 phase is consistently larger than that of tetrataenite. A contraction of the unit cell volume upon chemical ordering may be of fundamental importance regarding documented changes in the magnetic properties. Consistent with a symmetry reduction from cubic to tetragonal (A1deformation and annealing, a deformed FeNi(Ti) sample exhibits a decrease in the saturation magnetization = 10 K) of ~6% and an increase in the magnetic anisotropy upon annealing, Figure 80. These changes in the magnetic properties upon formation of the A6 phase are in agreement with observation of a ~5% decrease in the room temperature of bulk CoPt upon achievement of a transient state from a parent A1 phase, reported by Orehotsky [204,206]. They are also consistent with detected changes in saturation magnetization and magnetic anisotropy of meteoritic FeNi during A1 chemical disordering transformation [207], Figure 81. Furthermore, theoretical calcu

lations for ferromagnetic alloys which e
lations for ferromagnetic alloys which exhibit an fcc-based chemically ordered phase have shown that a small enhancement in the magnetocrystalline anisotropy can be expected from the development of tetragonality alone, while compositional ordering results in a much larger effect [138]. Thus, these results corroborate the d FeNi-based materials of this study. r why the A1 phase would transform into a tetragonal A6 phase. Tetragonal distortions present in near-equiatomic -forming systems have been, to date, associated with the development of long-range order. However, experimental evidence presented in this dissertation shows that tetragonality is achieved in the studied FeNi samples without long-range order, although a certain extent of short-rarmore, it was shown that the formation of the A6 phase in deformed FeNi samples was a result of annealing, implying a thermally-activated process. This feature suggests that the A6 phase is not an unstable phase with a transitory existence resulting from severe plastic deformation, but that it is in fact an equilibrium phase in the FeNi system. Under such circumstances, two possible thermodynamic scenarios for phase transformations in near-equiatomic FeNi are proposed. In the first case, there might exist a very small chemical driving force () that, coupled with the low atom mobility in this system at low temperatures, limits the rate of the A1A6 transformation. Severe

plastic deformation can modify the free
plastic deformation can modify the free energy of the A1 phase (Deformed A1), resulting in an increased driving force for the transition to the A6 phase. This scenario would imply that undeformed A1 FeNi annealed for a sufficiently long time may eventually transform into the A6 phase. Alternatively, the chemical free energy of the A1 phase may be lower than that of the A6 phase () and an increase in G as a result of plastic deformation provides the driving A6 transition (i.e. G Deformed A1). In this case, undeformed A1 would never transform to the tetragonal A6 phase. Consideration of reports of phase relationships in meteorites can provide insight concerning which of these scenarios most accurately describes the thermodynamic aspects of near-equiatomic FeNi. In meteorites, phase forms under proper time and temperature conditions without prior deformation, implying that GL1o below the equilibrium temperature evidence for the existence of an A6 phase in meteorite-derived samples has not been reported to date, although Bordeaux demonstrated the presence of an intermediate magnetic state between the A1 and L1 phases in the NWA6259 meteorite [207], which is hypothesized here to correspond to the A6 phase. If this is the case then L1o, and the first scenario described here is a better representation of thermodynamic aspects of near-equiatomic FeNi, wherein severe plastic deformation increases the driving force for

the A1A6 transition. This explanation is
the A1A6 transition. This explanation is in agreement with reports of attainment of an A6-type symmetry in the closely related FePd system, which is reported to link the A1 and L1 phases [46]. Experimental investigations aimed at determining whether the A6 phase can be produced without the use of plastic deformation may be able to validate this hypothesis, as proposed in Section 6 - of this document. A proposed sequence of transformation in FeNi, starting from a parent cubic A1 phase to produce a tetragonal chemically ordered L1 phase, is a two-step mechanism where a tetragonal distortion of the lattice is accommodated first during the A1A6 transformation, followed by nucleation and subsequent growth of the L1 phase ), Figure 82. The tetragonal distortion characteristic of the first step of the transformation (A1A6) suggests an initial displacive ( diffusionless) transition, while the second step of the transformation (A6) proceeds through a diffusional vacancy-controlled mechanism. It is still an open question what is the degree of short-range order in the A6 phase, and if it is any different to that present in the parent A1 phase. If the short-range orA6 step of the transformation, then that step might actually involve short-range diffusion. Considering the close structural relationships between the A1, A6, and L1 crystal structures, the presence of additional transitional states between the A1 and L1 phases

with a different crystal symmetry is con
with a different crystal symmetry is considered here as highly unlikely. Plastic deformation may not only modify the driving force for the formation of the A6 phase, but may also lower the energy barrier. This proposed two-step mechanism would indicate that formation of the A6 phase by severe plastic deformation followets a positive step towards achievement of L1Figure 82. Proposed sequence of phase transformation in Fe, where A1 is the parent cubic fcc phase, A6 is an intermediate tetragonal phase, and L1 is the product tetragonal chemically ordered structureThe influence of ternary alloying additions on the structural character of FeNi is presented in this section. Additions of Ti, V, and Al (2 at%) to near-equiatomic FeNi provide an increase in the unit cell volume relative to an unmodified FeNi counterpart with the same Fe content, Figure 73 and Figure 78. The tetragonal A6 phase, detected in processed FeNi(Ti) and FeNi samples, was not present in FeNi(Al), and FeNi(V) compositions within the limits of detection. However, it is unclear if Al and V additions to FeNi do not favor the formation of the A6 phase, or if the tetragonality developed in these systems is more subtle and thus more challenging to detect. It is anticipated that future neutron diffraction experiments performed on these samples can provide more information regarding the presence of the A6 phase. The tetragonality of the A6 phase in FeN

i samples containing Ti is consistently
i samples containing Ti is consistently higher than in unmodified binary FeNi, Figure 77. Additions of Ti also expand the unit cell lattice relative to the FeNi samples, providing a relatively larger unit cell volume, Figure 78. This indicates that the proper ternary additions to near-equiatomic FeNi may favor the tetragonality in the FeNi system, with implications regarding the magnetic character. However, it is unclear at the present time if Ti additions had an effect on the thermodynamic aspects of the A1 transformation in FeNi. Significance of work on composition-processing-structure-property correlations in FeNi The structural and magnetic results presented in this Section provide compelling evidence of a processing-induced tetragonality in bulk near-equiatomic FeNi-based samples. This result constitutes the creation of a new chemically disordered tetragonal phase in the Fe-Ni system. Severe plastic deformation in these alloys followed by annealing results in a tetragonal distortion of ~0.3% with no change in unit cell volume and no long-range chemical order. This tetragonal chemically disordered state, unreported to date, is identified as an A6 phase, and is proposed as an intermediate phase between the cubic parent A1 phase and the tetragonal chemically ordered L1 phase. It exhibits modified magnetic properties relative to the parent phase, with an ~6% lower and a slight increase in magnetocrystalline a

nisotropy. These conclusions are consist
nisotropy. These conclusions are consistent with reports of a transitional tetragonal state in the related ferromagnetic FePd [45,46,207] and CoPt [204–206] systems, with modified intrinsic magnetic properties in relation to both the A1 and L1The identification of a new phase in bulk near-equiatomic FeNi-based samples, reliably produced by severe plastic deformaknowledge concerning the Fe-Ni binary system, for which the highly sluggish kinetics have hindered the study of phase transformations and equilibrium phases at low temperatures. Furthermore, achievement of this intermediate tetragonal A6 phase indicates positive progress towards the laboratory synthesis of bulk L1 FeNi, with promise for the realization of next-generation advanced permanent magnets. The primary goal of this dissertation is to understand chemical ordering in near-equiatomic FePd and FeNi systems as a function of intrinsic and extrinsic parameter variation. This goal was pursued by investigating three different but correlated aspects: the effects of ternary alloying additions on the structure, magnetism and phase transformation character of near-equiatomic FePd, ) the effects of severe plastic deformation on the A1 transformation of near-equiatomic FePd, and of ternary alloying additions and severe plastic deformation on the A1transformation in near-equiatomic FeNi. Comparison of results obtained from unmodified FePd/FeNi alloys with thos

e of their chemically/microstructurally-
e of their chemically/microstructurally-modified counterparts is demonstrated to provide quantitative insight into fundamental composition-structure-le to advance the understanding of L1chemical ordering in Fe-based compounds. This Chapter presents a summary of the The effects of ternary alloying additions of Ni or Cu on the structure, magnetism and phase transformation character of near-equiatomic FePd were reported in Section 4.1. It was demonstrated that the chemical ordering transformation is very sensitive to compositional modification by incorporation of a ternary element, and that Ni and Cu additions affect differently the thermodynamic and kinetic aspects of the transition, as well as the resultant characteristics of the L1 phase. It is found that additions of Ni to near-equiatomic FePd have a profound effect on the chemical order-disorder temperature and on the overall enthalpy of transformation, decreasing the driving force for the transition and resulting in overall sluggish kinetics in the ternary FePdNi system. The isothermal 773 K section of the ternary Fe-Pd-Ni phase diagram near the equiatomic composition was obtained from analyses of the experimental results obtained in this dissertation. The Fe-Pd-Ni phase diagram displays a very narrow L1stability 48.251.8- compositions allows of Ni solubility. Furthermore, a narrow fcc + L1 phase field was determined for 6 at% at this temperature. The str

uctural characteristics of the L1 FePd p
uctural characteristics of the L1 FePd phase are also affected by Ni additions, with the unit cell volume decreasing with increased Ni content. In contrast to the effects of Ni additions, additions of Cu to near-equiatomic FePd are determined to have a more subtle effect on of the A1 transformationAs a result, a wider range of L1 phase stability at 773 K is projected for the Fe-Pd-Cu ternary system, which is proposed to allow = 18 at% solubility of Cu (or more) for Fecompositions. The isothermal 773 K section of the ternary Fe-Pd-Cu phase diagram near the equiatomic composition was defined from the experimental results obtained in this dissertation and from reports by other authors, displaying an equilibrium phase field for Fe50-x ternary alloys with Cu content 16 at%. Additions of Cu into the FePd lattice significantly reduce the axial ratio, without an impact on the unit cell volume. The effects of chemical modification by Ni/Cu ternary alloying additions on the Curie temperature of fcc and L1 FePd phases were compared in terms of the average valence electron concentration of the alloys. It was demonstrated that of fcc and L1phases in ternary FePdM with varying M content decreases systematically with increased valence electron concentration, reproducing the reported behavior of for fcc and L1phases in binary FePd with varying Fe-content. This relation can be used to design of near-equiatomic fcc and L

1 FePd phases through the addition of 3e
1 FePd phases through the addition of 3elements other than those studied here. These conclusions obtained from studies of the model FePd system can be extrapolated to the FeNi system, for whadditions can be used to tune the thermodynamic and kinetic aspects of the A1transformation. Significantly, it is predicted that ternary alloying additions to near-equiatomic FeNi that increase can favor the A1 transformation by either amplifying the driving force for nucleation of the L1 phase or by allohigher temperatures to foster increased lattice diffusivities. The effects of severe plastic deformation delivered through cold-rolling on the transformation of near-equiatomic FePd were analyzed in Section 4.2. Application of strain to the near-equiatomic L1-type FePd produces a crystallographic texture, reduces the crystallite size, and destroys the long-range chemical order. However, under the processing conditions studied here, complete chemical disordering is not attained. The mechanism by which long-range order is lost is purely mechanical, related to dislocation generation and motion that quickly, but not completely, transforms structure into a chemically disordered phase. The chemical ordering A1transformation was investigated in undeformed and deformed near-equiatomic FePd through isothermal DSC measurements, producing data that was analyzed using a JMAK model. In this manner, fundamental differences of the nucle

ation and growth mechanisms between unst
ation and growth mechanisms between unstrained and strained samples were identified. It was determined that the transformation in undeformed bulk FePd is initiated by a burst of nucleation of phase followed by a two-dimensional interface controlled growth. The effective activation energy for this process, determined here for the first time as 2.94 ± 0.48 eV/atom, is comparable to the reported activation energy for lattice diffusion in this system. The agreement between both activation energies suggests that the rate of the transformation in undeformed FePd is likely limited by the requirement to create and move vacancies to the interface between transformed and untransformed regions. In comparison, L1 growth in deformed samples is determined to occur by a one-dimensional mechanism. It is suggested that this difference between the growth mechanism found for strained versus unstrained systems is related to the increased amount of grain boundaries and other structural defects induced in the deformed samples, which facilitate heterogeneous nucleation of the L1 phase. The nuclei in the deformed sample then coalesce and grow as a planar front. The effective activation energy for chemical ordering in deformed FePd was calculated here for the first time as 1.96 ± 0.21 eV/atom, much lower than the reported activation energy for diffusion. It is suggested that the highly defective microstructure of deformed FePd facil

itates vacancy formation at the interfac
itates vacancy formation at the interface between transformed and untransformed regions, enhancing the kinetics of the transformation in comparison with that of undeformed samples. Conclusions obtained here from results of the model FePd system can be extrapolated to the FeNi system, where severe plastic deformation prior to annealing is anticipated to favor the transformation. The effects of ternary alloying additions (Ti, V, Al) and severe plastic deformation (cold-rolling/cryomilling) on the A1 transformation in near-equiatomic FeNi were examined in Section 4.3. A tetragonal chemically disordered A6 structure, unreported to date for this system, was detected in FeNi and FeNi(Ti) alloys subjected to deformation and subsequent annealing. It was confirmed that the onset of tetragonality does not occur until annealing is applied. This tetragonal A6 phase has the same unit cell volume as the original A1 phase, but exhibits a tetragonal distortion of ~0.3%, which is in agreement with that reported for the tetragonal L1 FeNi phase in meteorite-derived samples. However, it was unequivocally confirmed that this A6 phase does not exhibit long-range chemical order. Nonetheless, the detection of the presence of local structural correlations in the deformed, annealed FeNi specimens suggests that the A6 phase possesses a certain degree of short-range order. The A6 FeNi phase has modified magnetic properties relative to

those of the A1 FeNi parent phase: consi
those of the A1 FeNi parent phase: consistent with the symmetry reduction from cubic to tetragonal, a decrease in the of ~6% and an increase in the magnetocrystalline anisotropy is documented in a deformed FeNi-based sample after annealing. Overall, this newly discovered A6 FeNi phase is envisioned as a transitional state between the A1 and L1 phases in FeNi, as is reported for FePd. These results, together with those reported by other authors on the L1-forming FePd and CoPt systems, indicate that the transformation from a parent cubic A1 phase to a tetragonal chemically ordered L1 phase in ferromagnetic systems may occur a two-step mechanism: an initial tetragonal distortion of the lattice (A1A6) is followed by nucleation and growth of the chemically ordered phase (A6vacancy-controlled mechanism. In FeNi, severe plastic deformation followed by of the phase transformation (A1A6), and processing techniques to accelerate the chemical ordering process still need to be identified. The A6 phase was not detected in FeNi alloys modified with Al or V additions. However, the A6 phase was detected in alloys modified with Ti additions, with larger ratios with respect to unmodified alloys, indicating Ti additions favor the development of Overall, the results obtained in this dissertation furnish new knowledge concerning the Fe-Ni material system, and advance the understanding of L1 chemical RECOMMENDATIONS FOR FUTURE

WORK In this dissertation, the influence
WORK In this dissertation, the influence of intrinsic and extrinsic modification on the phase transformation in FePd and FeNi-based system was studied. Composition-structure-processing-property correlations were analyzed, and insight was gained regarding the mechanism of the L1 chemical ordering transformation in ferromagnetic systems. This work opens up several pathways for future work, aimed at addressing important questions that can advance further the understanding of the transformations to eventually achieve the laboratory synthesis of L1 FeNi. Examples of these questions are provided below, and explained in more detail later as part of the recommendations for future work: Are there elements that can increase the in FeNi? Can a model with predictive capabilities be obtained to determine the effect of ternary alloying additions on of the A1 transformation in ferromagnetic systems? In this manner, proper processing windows for the attainment of L1 FeNi can be identified. Which processing techniques can enhance low temperature atomic diffusion in the FeNi system? These techniques can be evaluated to promote the formation of the L1rmanent magnet applications. What are the strain/annealing-time threshold limits for attainment of the A6 FeNi phase? Can the A6-type FeNi phase be obtained without application of strain through annealing alone, and if so, for how long should the thermal treatment be made to obta

in the same results produced by deformat
in the same results produced by deformation? This can advance the understanding of the relation between strain/temperature and the A6 phase formation in ferromagnetic systems. Recommendations for future research avenues to address the questions above are presented in this Section. It is anticipated that the experiments proposed will provide an enhanced understanding of the physics and mechanisms underlying the A1 phase transformation in ferromagnetic systems, ultimately providing relevant information for Recommendation 1Examination of additional ternary alloying additions on the In this dissertation, effects of ternary alloying additions of Ni, Cu / Ti, V, Al on transformation in FePd / FeNi systems were studied, revealing important differences in the effects of the different ternary elements. The results obtained here on the model FePd system advance the suggestion that the effect of element on the chemical order-disorder temperature is the main factor that limits the formation of an L1 phase, but it is still unclear how the identity of the ternary element . To clarify these aspects, further experimental work on ternary FePdM and FeNiM systems should focus on: Studying the behavior of FePd alloys subjected to the addition of other -elements covering different groups in the periodic table, to obtain more information on the variation of with composition. This will allow to draw relations that eventually can

serve to predict the effect of ternary
serve to predict the effect of ternary addition Studying other ternary additions to FeNi to evaluate their potential for the formation of the L1 phases. Since it is unknown which elements may serve to promote the formation of these phases, a useful starting point would be to select elements that form thermodynamically stable L1 phases containing either Fe or Ni. These include ong others. Ideally, the ternary element of FeNi, to favor the formation of the L1Recommendation 2:Investigate other severe plastic deformation techniques as phase formation in FeNi The results of this dissertation demonstrate that severe plastic deformation applied to the model near-equiatomic FePd system has a profound impact on the operative nucleation and growth mechanisms and on the effective activation energy of the transformation, resulting in overall enhanced kinetics in strained samples as compared to those found in the unstrained alloys. Furthermore, severe plastic deformation applied to near-equiatomic FeNi-based alloys reliably produced a tetragonal distortion of the cubic lattice, producing the tetragonal chemically disordered A6 crystal structure, but did not seem to be successful in accelerating the chemical ordering process to attain the L1 phase. The development of chemical order in FeNi is critical for the attainment of high magnetocrystalline anisotropy, a necessary requirement for permanent magnet applications. Ther

efore, processing techniques that enhanc
efore, processing techniques that enhance diffusion at low temperatures in the FeNi system to promote chemical ordering in the tetragonal A6 phase still need to be identified. One avenue of future research should explore a spectrum of severe plastic deformation techniques that can achieve much higher strain levels than those studied here. For instance, high pressure torsion is a severe plastic deformation technique known to deliver extremely high strain [190,191]. Other possibility that may be explored is the application of strain simultaneous to the annealing process in a furnace, known as “strain-annealing”. This technique has been used in Fe and Co-based soft magnetic materials for power conversion, in order to induce magnetic anisotropy and reduce energy losses [208–212]. When applied to soft-magnetic alloys, the technique makes use of a ribbon of the material in an initial amorphous state. The ribbon is subjected to tension while traversing a furnace, as displayed in Figure 83. Strain-annealing induces crystallization of the ribbons and an easy-direction of magnetization along the Figure 83. Schematic representation of strain annealing procedure. Image g procedure. Image . It is anticipated that simultaneous application of stress and temperature in the FeNi system would result in the simultaneous formation of a high density of structural defects that serve as diffusion pathways, and the immediate use of t

hese defects for the necessary atomic re
hese defects for the necessary atomic rearrangement to produce the chemically ordered L1 phase. The initial FeNi ribbons need not be in an amorphous state as is the case when strain-annealing is applied to soft-magnetic materials, although an amorphous phase could in fact benefit diffusion processes at low temperatures. However, production of amorphous FeNi alloys would require incorporation of glass-forming elements [214,215], which will have an effect on the final structural and magnetic character of the alloy. Recommendation 3. Further investigate the intermediate A6 phase in the FeNi transformation The results described in Section 4.3 of this dissertation represent the first report of a tetragonal, chemically disordered A6 phase in near-equiatomic FeNi alloys. In the study of A1 chemical ordering transformations in ferrous systems, it has generally been assumed that the origin of the tetragonal distortion experienced upon formation of the L1phase is directly related to the development of long-range order [216–218]. In fact, the c/a ratio is often used as an indication of the degree of long-range order, alluding to the correspondence between the alluding to the correspondence between the nce of a tetragonal but chemically disordered phase in FeNi reported in this dissertation, together with limited reports of this phase in other L1-forming alloys, implies that under certain conditions the development of t

etragonality and of chemical order may n
etragonality and of chemical order may not be concurrent. It is anticipated that computational studies and further experimental investigations on the A6 phase will be useful for developing a deeper understanding of the mechanism of L1chemical ordering in ferromagnetic systems, of interest from both a fundamental science point of view as well as for the design and optimization of L1-based magnets. Some recommendations for further experimentation include: Structural examination of undeformed annealed FeNi samples – In this dissertation, the A6 FeNi phase was consistently obtained through severe plastic deformation followed by annealing. To better understand thplastic deformation and the formation of the A6 phase, it is of interest to identify if this phase can be accomplished by annealing alone, and if it can, for how long does the thermal treatment need to be to reproduce what plastic deformations does. These studies will need to make use of high-resolution structural characterization techniques as the ones used in this dissertation. To advance the understanding of the nature of the A6 phase in FeNi, it is of interest to characterize the strain/annealing-time threshold limits for attainment of the A6 phase. Examination of different FeNi-based samples processed with significantly different levels of strain and for different annealing times is proposed for future experiments. Here, use of high-resolution structur

al characterization techniques would be
al characterization techniques would be of critical importance for characterization of Identification of the A6 and L1 phases in FeNi has proved to be extremely challenging. In this dissertation, high-resolution structural probes were used to characterize the FeNi-based alloys, but the use of these techniques requires use of national user research facilities implying significant investment in terms of time and resources. In-house techniques capasignature of the L1 or A6 phase should be sought. Magnetic probes are challenging because the A6 FeNi magnetic signature is very subtle, and an L1 FeNi magnetic signal requires large volume fractions of the L1 phase aligned in a preferential direction. As was seen in Section 2.1, Mössbauer spectroscopy as well as mechanical and electric measurements can be used to identify signatures of chemical ordering. It is proposed that future studies evaluate the potential of these techniques to identify Unit cell lattice constants Miller indices Fractional coordinates of atom in unit cell Terms in the Law of Approach to Saturation. relates to the existence of structural defects/magnetic inclusions. relates to magnetocrystalline A1 fcc chemically disordered phase A2 bcc chemically disordered phase A6 fct chemically disordered phase Cross sectional area of Measured full-width at half-maximum for the sample Isotropic temperature factor Energy density - figure of merit c Speed of

light eation sites per unit volume conta
light eation sites per unit volume contained Heat capacity at constant pressure, at constant magnetic field Volume fraction of Interplanar distance Charge of electron Elastic modulus planes Exchange interaction Scattering factor of atom Structure factor Gibbs free energy Gibbs free energy of initial state, final state Infinite thickness in XRD 0, Thickness of the material: initial, final h(t) Heat flow as a function of time Applied magnetic field Internal field Coercivity Intensity of incident X-ray beambeam Theoretical integrated intensity of a Bragg peak. Subscript indicates a phase Experimental integrated intensity of a Bragg peak. Subscript indicates a Boltzmann constant (T) Temperature dependent rate constant Magnetic moment Mass of electron Remanence Saturation magnetization Avrami exponent Multiplicity factor for Bragg reflection phase, fraction of -sites occupied by A atoms, fraction of -sites occupied by B atoms Ideal gas constant Radius of diffractometer Electron spin of atoms and Thermal entropy Time Initial time of phase transformation, final time of phase transformation Temperature Annealing temperature Curie temperature Critical chemical order-disorder temperature Internal energy Volume of unit cell Volume phase, atomic fraction of A species, atomic fraction of B species X(t) Fraction transformed during a phase transformation Fraction transformed during a phase transformation measured fro

m DSC Compressibility Sample contributi
m DSC Compressibility Sample contribution to peak breadth Magnetic susceptibility, magnetic susceptant temperature Anomalous scattering factor Nucleation barrier Free energy associated with destruction of defects Activation energy for atomic migration Elastic strain energy associated with misfit of product phase in parent phase Free energy associated with formati Undercooling Microstrain phase, fraction of -sites occupied by A atoms, fraction of -sites occupied by B atoms Instrumental contribution to peak breadth of the monochromator and crystal Radiation wavelength Linear absorption coefficient Factor that includes vibroms and the area of the Angle between easy direction and magnetization vector Electrical resistivity Uniform stress Strength Incubation time Bragg angle of the monochromator Bragg angle of the analyzer crystal APB Anti-phase boundaries bcc Body-centered cubic bct Body-centered tetragonal BNL Brookhaven National Lab DSC Differential scanning calorimetry EDS Energy dispersive X-ray spectroscopy fcc Face-centered cubic fct Face-centered tetragonal FWHM Full-width at half-maximum JMAK Johnson-Mehl-Avrami-Kolmogorov LRO Long-range order NSLS National Synchrotron Light Source SAED Selected area electron diffraction SEM Scanning electron microscopy SQUID Superconducting quantum interference device STA Simultaneous thermal analyzer TEM Transmission electron microscopy TG Thermogravimetry TTT

Transformation-temperature-time VSM Vi
Transformation-temperature-time VSM Vibrating sample magnetometer XRD X-ray diffraction XRF X-ray fluorescence %CW Percentage of cold work [1] J.F. Herbst, RB materials: Intrinsic properties and technological aspects, [2] B.K.J. Strnat, Modern Permanent Magnets for Applications in Electro - Technology, Proceedings of the IEEE. 78 (1990) 1–35. doi:10.1109/5.56908. [3] N. Poudyal, J. Ping Liu, Advances in nanostructured permanent magnets research, Journal of Physics D: Applied Physics. 46 (2013) 043001. doi:10.1088/0022-[4] O. Gutfleisch, M. a Willard, E. Brück, C.H. Chen, S.G. Sankar, J.P. Liu, Magnetic materials and devices for the 21st century: stronger, lighter, and more energy efficient., Advanced Materials (Deerfield Beach, Fla.). 23 (2011) 821–42. doi:10.1002/adma.201002180. [5] K. Smith Stegen, Heavy rare earths, permanent magnets, and renewable energies: An imminent crisis, Energy Policy. 79 (2015) 1–8. [6] J.M.D. Coey, Hard Magnetic Materials: A Perspective, IEEE Transactions on [7] R. Skomski, J.M.D. Coey, Magnetic anisotropy — How much is enough for a permanent magnet?, Scripta Materialia. 112 (2016) 3–8. doi:10.1016/j.scriptamat.2015.09.021. [8] L.H. Lewis, F. Jimenez-Villacorta, Perspectives on Permanent Magnetic Materials tion, Metallurgical and Materials [9] R. Stamps, S. Breitkreutz, J. Åkerman, A. Chumak, Y. Otani, G. Bauer, et al., The 2014 Magnetism Roadmap, Journal of Applied Physics D: Ap

plied Physics. 47 [10] T. Klemmer, D. Ho
plied Physics. 47 [10] T. Klemmer, D. Hoydick, H. Okumura, B. Zhang, W.A. Soffa, Magnetic hardening and coercivity mechanism in L1 ordered FePd ferromagnets, Scripta [11] A. Hirohata, H. Sukegawa, H. Yanagihara, I. Zutic, T. Seki, S. Mizukami, et al., Roadmap for emerging materials for spintronic devices, Condensed Matter. [12] R.W. McCallum, L.H. Lewis, R. Skomski, M.J. Kramer, I.E. Anderson, Practical Aspects of Modern and Future Permanent Magnets, Annu. Rev. Mater. Res. 44 nnurev-matsci-070813-113457. [13] L.H. Lewis, F. Jiménez-Villacorta, Perspectives on Permanent Magnetic Materials tion, Metallurgical and Materials [14] L. Néel, J. Pauleve, R. Pauthenet, J. Laugier, D. Dautreppe, an Iron—Nickel single crystal ordered by neutron bombardment, Journal of [15] J. Paulevé, A. Chamberod, K. Krebs, A. Bourret, Magnetization curves of FeNi (50–50) single crystals ordered by neutron irradiation with an applied magnetic [16] K.B. Reuter, D.B. Williams, J.I. Goldstein, Ordering in the Fe Ni system under [17] W. Chambron, D. Dautreppe, L. Néel, J. Paulevé, Irradiation aux électrons de 1 MeV d’un alliage Fe-Ni (50-50%), Comptes Rendus de L’Académie Des Sciences. [18] T. Shima, M. Okamura, S. Mitani, K. Takanashi, Structure and magnetic -ordered FeNi films prepared by alternate monatomic layer deposition, Journal of Magnetism and 310 (2007) 2213–2214. doi:10.1016/j.jmmm.2006.10.799. [19] T. Kojima, M. Mizuguchi, K. Ta

kanashi, L1-ordered FeNi film grown on C
kanashi, L1-ordered FeNi film grown on Cu-Ni binary buffer layer, Journal of Physics: Conference Series. 266 (2011). [20] T. Kojima, M. Ogiwara, M. Mizuguchi, M. Kotsugi, T. Koganezawa, T. Ohtsuki, et al., Fe-Ni composition dependence of magnetic anisotropy in artificially fabricated L1-ordered FeNi films, Journal of Physics: Condensed Matter. 26 [21] T. Kojima, M. Mizuguchi, T. Koganezawa, M. Ogiwara, M. Kotsugi, T. Ohtsuki, et al., Addition of Co to L1-ordered FeNi films: influences on magnetic properties [22] T. Kojima, M. Ogiwara, M. Mizuguchi, M. Kotsugi, T. Koganezawa, T. Ohtsuki, et al., Fe-Ni composition dependence of magnetic anisotropy in artificially fabricated L1-ordered FeNi films., Journal of Physics. Condensed Matter. 26 [23] B. Zhang, M. Lelovic, W.A. Soffa, The formation of polytwinned structures in Fe- Pt and Fe- Pd alloys, Scripta Meta[24] D.E. Laughlin, K. Srinivasan, M. Tanase, L. Wang, Crystallographic aspects of magnetic materials, Scripta Materialia. 53 (2005) 383–388. doi:10.1016/j.scriptamat.2005.04.039. [25] A.K. Kulovits, A.R. Deshpande, J.M.K. Wiezorek, Cold-working and annealing of the chemically ordered L1-phase Fe, Intermetallics. 17 (2009) 865–871. doi:10.1016/j.intermet.2009.03.015. [26] H. Okamoto, Phase diagrams of binary alloys, ASM International, Materials Park, OH, 1993. [27] D. Porter, K. Easterling, M. Sherif, Phase transformations in metals and alloys, 3rd Editio, CRC P

ress Taylor & Franci[28] F. Nix, W. Shoc
ress Taylor & Franci[28] F. Nix, W. Shockley, Order-disorder transformations in alloys, Reviews of Modern Physics. 10 (1938). http://rmp.aps.org/abstract/RMP/v10/i1/p1_1 (accessed July [29] O. Dahl, Cold-work deformation and recovery in alloys with orderly atom configuration, Zeitschrift [30] K. Barmak, K. Coffey, Metallic films for electronic, optical and magnetic d properties, Woodhead Publishing, 2014. [31] J.B. Newkirk, A.H. Geisler, D.L. Martin, R. Smoluchowski, Ordering reaction in cobalt-platinum alloys, Jour[32] H. Baker, H. Okamoto, ASM Handbook Volume 03 - Alloys Phase Diagrams, [33] A. Chbihi, X. Sauvage, C. Genevois, D. Blavette, D. Gunderov, A.G. Popov, Disordering and Ordering in a Severely Deformed FePd Alloy, Solid State Phenomena. 172-174 (2011) 703–708. [34] B. Cheong, D. Laughlin, Morphology of structural domains in a congruently phase FePd alloy, Scripta Metallurgica et Materialia. 29 (1993) 829–[35] A. R. Deshpande, J.M.K. Wiezorek, Magnetic age hardening of cold-deformed bulk equiatomic Fe–Pd intermetallics during isothermal annealing, Journal of Magnetism and Magnetic Materials. 270 (2004) 157–166. doi:10.1016/j.jmmm.2003.08.013. [36] A. R. Deshpande, J.R. Blachere, J.M.K. Wiezorek, Texture evolution in combined reaction transformed equiatomic ferromagnetic L1-ordered FePd intermetallics, i:10.1016/j.scriptamat.2005.10.026. [37] B. Zhang, W.A. Soffa, Structure and properties of raand i

ron-palladium alloys, IEEE Transactions
ron-palladium alloys, IEEE Transactions on Magnetics. 26 (1990) 1388–[38] L. Wang, Z. Fan, A.G. Roy, D.E. Laughlin, Effects of atomic ordering on the Curie temperature of FePd L10 type alloys, Journal of Applied Physics. 95 (2004) [39] G.M. Gushchin, F.N. Berseneva, Microstructure and ordering kinetics of quenched [40] N.C. Bordeaux, Structure-magnetism correlations and chemical order-disorder transformations in ferrous L1-structured compounds, Northeastern University, [41] B. Zhang, W. Soffa, Magnetic domains and coercivity in polytwinned ferromagnets, Physica Status So[42] B. Zhang, W. a. Soffa, Microstructure and magnetic domain structure in Fe-Pt and ctions on Magnetics. 27 (1991) 5328–[43] T.J. Klemmer, C. Liu, N. Shukla, X.W. Wu, D. Weller, M. Tanase, et al., Combined reactions associated with L1 ordering, Journal of Magnetism and [44] N.I. Vlasova, A.G. Popov, N.N. Shchegoleva, Effect of the structural state of the FePd equiatomic alloy on the temperature dependence of the initial magnetic susceptibility and the curie temperature, The Physics of Metals and Metallography. 107 (2009) 359–369. doi:10.1134/S0031918X09040061. [45] N.I. Vlasova, N.N. Shchegoleva, A.G. Popov, G.S. Kandaurova, Ferroelastic Domains and Phases in Ferromagnetic Nanostructured FePd Alloy, The Physics of Metals and Metallography. 110 (2010) 449–463. [46] N.I. Vlasova, a. G. Popov, N.N. Shchegoleva, V.S. Gaviko, L. a. Stashkova, G.S.

Kandaurova, et al., Discovery of metast
Kandaurova, et al., Discovery of metastable tetragonal disordered phase upon phase transitions in the equiatomic nanostructured FePd alloy, Acta Materialia. 61 (2013) 2560–2570. doi:10.1016/j.actamat.2013.01.033. [47] A. Cebollada, R.F.C. Farrow, M.F. Toney, Structure and magnetic properties of chemically ordered magnetic binary alloys in thin film form, in: H.S. Nalwa (Ed.), Magnetic Nanostructures, American Scientific Publishers, Stevenson Ranch, [48] R. Hultgren, C.A. Zapffe, The crystal structures of the iron-palladium superlattices, Zeitschrift Fuer Kristallographie, KristallgeomeKristallchemie. 99 (1938) 509–12. [49] E. Raub, H. Beeskow, O. Loebich Jr, The phase diagram Iron-Palladium below [50] V.A. Kussmann, K. Muller, Magnetische Eigenschaften der geordneten phase FePd im homogenen und heterogenen bereich, Zeitschrift Fur Angewandte Physik. [51] H. Shima, K. Oikawa, A. Fujita, K. Fukamichi, K. Ishida, A. Sakuma, Lattice axial ratio and large uniaxial magnetocrystalline anisotropy in L1-type FePd single crystals prepared under compressive stress, Physical Review B. 70 (2004) 224408. [52] C.W. Yang, D.B. Williams, J.I. Goldstein, A revision of the Fe-Ni phase diagram at low temperatures (00 C), Journal of Phase Equilibria. 17 (1996) 522–531. [53] K.B. Reuter, D.B. Willaims, J.I. Goldstein, Determination of the Fe-Ni Phase Diagram below 400 C, Metall and Mat Trans A. 20 (1989) 719–725. [54] R. Wakelin, E.

Yates, A study of the order-disorder tr
Yates, A study of the order-disorder transformation in Iron-Nickel alloys in the region FeNi, Proceedings of the Physical Society. Section B. 66 [55] J. Paulevé, D. D, Orientational superlattice of a Fe-Ni (50-neutron irradiation in a magnetic field, Comptes Rendus de L’Académie Des [56] J. Paulevé, D. Dautreppe, J. Laugier, L. Néel, A new order-disorder transition in [57] J. Petersen, M. Aydin, J. Knudsen, Mössbauer spectroscopy of an ordered phase (superstructure) of FeNi in an iron meteorite, Physics Letters A. 62A (1977) 192–[58] R.S. Clarke Jr, E.R.D. Scott, Tetrataenite - ordered FeNi, a new mineral in meteorites, American Mine[59] J.F. Albertsen, M. Aydin, J.M. Knudsen, Mossbauer Effect Studies of Taenite Lamellae of an Iron Meteorite Cape York [60] J.I. Goldstein, J.M. Short, The iron meteorites, their thermal history and parent bodies, Geochimica et Cosmochimica Acta. 31 (1967) 1733–1770. [61] C.W. Yang, D.B. Willaims, J.I. Goldstein, A new empirical cooling rate indicator for meteorites based on the size of the cloudy zone of the metallic phases, [62] J.F. Albertsen, Tetragonal lattice of tetrataenite (ordered Fe-Ni, 50-50) from 4 meteorites, Physica Scripta. 23 (1981) 301–306. [63] C.W. Yang, J.I. Goldstein, Phase decomposition of invar alloy - information from the study of meteorites, in: The Invar Effect: A Centennial Symposium, The [64] N. Bordeaux, A.M. Montes-Arango, J. Liu, K. Barmak, L.H. Lew

is, Thermodynamic and Kinetic Parameters
is, Thermodynamic and Kinetic Parameters of the Chemical Order-Disorder Transformation in L1 FeNi (Tetrataenite), Acta Materialia. 103 (2016) 608. doi:10.1016/j.actamat.2015.10.042. [65] J. Paulevé, D. Dautreppe, J. Laugier, L. Néel, Establishment of an ordered FeNi structure by neutron irradiation, Comptes Rendus de L’Académie Des Sciences. [66] J.W. Christian, The theory of transformations in metals and alloys, 3rd edition, [67] C. Zener, Theory of D for Atomic Diffusion in Metals, Journal of Applied [68] J.M. Penisson, A. Bourret, Mise en ordre de l’alliage Fe-Ni equiatomique sous ltage Electron Micros[69] J. Yang, J.I. Goldstein, Magnetic contribution to the interdiffusion coefficients in ) Fe-Ni alloys, Metallurgical and Materials Transactions A. 35 [70] Y. Geng, T. Ablekim, M.A. Koten, M. Weber, K. Lynn, J.E. Shield, Defect generation and analysis in mechanically alloyed stoichiometric Fe – Ni alloys, Journal of Alloys and Compounds. 633 (2015) 250–255. doi:10.1016/j.jallcom.2015.02.038. [71] S. Lee, K. Edalati, H. Iwaoka, Z. Horita, T. Ohtsuki, T. Ohkochi, et al., Formation of FeNi with L1-ordered structure using high-pressure torsion, Philosophical [72] M.C. Morris, H.F. McMurdie, E.H. Evans, B. Paretzkin, H.S. Parker, Panagiotopoulos, Standard X-ray Diffraction Powder Patterns, National Bureau of Standards, Monogr. 25 Section 18, US, 1981. [73] H. Swanson, H.F. McMurdie, M.C. Morris, E.H. Evans, Standa

rd X-ray Diffraction Powder Patterns, Na
rd X-ray Diffraction Powder Patterns, National Bureau of Standards, Monogr. 25 Section 5, US, 1967. [74] H.E. Swanson, M.I. Cook, E.H. Evans, J.H. DeGroot, Standard X-ray Diffraction Powder Patterns, National Bureau of Standards, Circular 539, Volume 10, 1960. [75] A. Makino, P. Sharma, K. Sato, A. Takeuchi, Y. Zhang, K. Takenaka, Artificially produced rare-earth free cosmic magnet, Scientific Reports. 5 (2015) 16627. [76] L.H. Lewis, F.E. Pinkerton, N. Bordeaux, A. Mubarok, E. Poirier, J.I. Goldstein, et al., De Magnete et Meteorite: Cosmically Motivated Materials, IEEE Magnetics [77] L.H. Lewis, A. Mubarok, E. Poirier, N. Bordeaux, P. Manchanda, A. Kashyap, et al., Inspired by nature: investigating tetrataenite for permanent magnet applications., Journal of Physics. Condensed Matter: An Institute of Physics [78] W. Xiong, H. Zhang, L. Vitos, M. Selleby, Magnetic phase diagram of the Fe–Ni system, Acta Materialia. 59 (2011) 521–530. doi:10.1016/j.actamat.2010.09.055. [79] A. Edström, J. Chico, A. Jakobsson, A. Bergman, J. Rusz, Electronic structure and magnetic properties of L1[80] D.E. Laughlin, Atomic and magnetic [81] B.D. Cullity, Elements of X-ray diffraction, Addison-Wesley Pub. Co., Reading, [82] J. Rodríguez-Carvajal, Recent developments in the program FULLPROF, Commission on Powder Di[83] J. Laugier, J. Paulevé, R. Pauthenet, Proprietés magnétiques d’un monocristal d'alliage Fe-Ni ordonné sous champ ma

gnétique par irradiation aux neutrons.,
gnétique par irradiation aux neutrons., Comptes Rendus de L’Académie[84] J.F. Albertsen, G.B. Jensen, J.M. Knudsen, Structure of taenite in two iron meteorites, Nature. 273 (1978) 453. [85] J. Albertsen, Tetragonal Lattice of Tetrataenite (Ordered Fe-Ni, 50-50) from 4 [86] J. Danon, R. Scorzelli, I. Azevedo, Iron–nickel 50–50 superstructure in the Santa [87] J. Danon, R. Scorzelli, I.S. Azevedo, W. Curvello, J.F. Albertsen, J.M. Knudsen, Mise en évidence par effet Mossbauer et diffraction de rayons X de la présence de l’alliage ordonné Fe-Ni 50-50 dans la méteorite Santa-Catharina, Comptes Rendus de L’Académie Des Sciences. 287 (1978) 199. [88] S. Mehta, P.M. Novotny, D.B. Williams, J.I. Goldstein, Electron-optical therville meteorite, [89] V.F. Sears, Neutron scattering lengths and cross sections, Neutron News. 3 (1992) [90] E. Kuzmann, S. Nagy, A. Vértes, Critical review of analytical applications of Mossbauer spectroscopy illustrated by mineralogical and geological examples, Pure and Applied Chemistry. 75 (2003) 801–858. [91] B. Fultz, Mössbauer Spectrometry, in: E. Kaufmann (Ed.), Characterization of Materials, John Wiley, New York, 2011. [92] L. Larsen, H. Roy-Poulsen, N.O. Roy-Poulsen, L. Vistisen, J.M. Knudsen, Order-Disorder Transitions in Iron-Nickel (50%-50%) Alloys from Iron Meteorites as Studied by Mössbauer Spectroscopy, Physical Review Letters. 48 (1982) 1054–[93] R. Gooley, R.B. Merril, J.R. Smyt

h, Anisotropic taenite in meteorites, Me
h, Anisotropic taenite in meteorites, Meteoritics. [94] R.S. Clarke, E.R.D. Scott, Tetrataenite--ordered FeNi, a new mineral in meteorites, American Mineralogist. 65 (1980) 624–630. [95] E.R.D. Scott, R.S. Clarke Jr, Identification of clear taenite in meteorites as ordered [96] A. Marchand, P. J., D. Dautreppe, Résistivité de l’alliage FeNi (50-50 %) ordonné par irradiation aux neutrons, Comptes Rendus de L’Académie Des Sciences. 257 [97] A. Marchand, A. Chamberod, Détermination des températures critiques d’ordre des alliages de fer-nickel, Comptes Rendus de L’Académie Des Sciences. 261 [98] T. Nagata, B. Carleton, Tetrataenite in chondritic meteorites, Proceedings of the [99] T. Nagatai, J.A. Danon, M. Funakii, Magnetic properties of Ni-rich iron meteorites, Memoirs of National Institute of Polar Research. Special Issue. 46 [100] C.L. Platt, K.W. Wierman, E.B. Svedberg, R. van de Veerdonk, J.K. Howard, a. G. Roy, et al., L1 ordering and microstructure of FePt thin films with Cu, Ag, and Au additive, Journal of Applied Physics. 92 (2002) 6104. doi:10.1063/1.1516870. [101] C.Y. You, Y.K. Takahashi, K. Hono, Particulate structure of FePt thin films enhanced by Au and Ag alloying, Journal of Applied Physics. 100 (2006) 056105. [102] T. Maeda, a. Kikitsu, T. Kai, T. Nagase, H. Aikawa, J. Akiyama, Effect of added Cu on disorder-order transformation of L1FePt, IEEE Transactions on Magnetics. [103] T. Kai, Magnetic a

nd electronic structures of FePtCu terna
nd electronic structures of FePtCu ternary ordered alloy, Journal [104] D.C. Berry, K. Barmak, Effect of alloy composition on the thermodynamic and kinetic parameters of the A1 to L1 transformation in FePt, FeNiPt, and FeCuPt films, Journal of Applied Physic[105] V.D. Sukhanov, V. V Maykov, E.I. Teytel, Structure and magnetic properties of ordered FePd-Me alloys (Me = Cr, V, Nb), Fizika Metallov I Metallovedenie. 77 [106] S. Shimizu, E. Hashimoto, Ordering in PtFe-PdFe and PtFe-PtNi alloys and their magnetic properties, J Japan Inst Metals. 35 (1971) 902. [107] H. Naganuma, K. Sato, Y. Hirotsu, Fabrication of oriented L1composite bcc-Fe-FeCuPd nanoparticles: Alloy composition dependence of magnetic properties, Journal of Applied Physics. 99 (2006) 08N706. [108] A. Kovács, K. Sato, Y. Hirotsu, Strong perpendicular magnetic anisotropy of Fe–Pd nanocrystalline particles enhanced by Co addition, Journal of Applied Physics. [109] J.-G. Kang, J.-G. Ha, Structure and magnetic properties of [110] T. Horiuchi, M. Igarashi, F. Abe, K. Ohkubo, S. Miura, T. Mohri, Phenomenological calculation of the Fe-Pd – Based L1-ordered phase in the Fe-Pd-Ni ternary system, Metallurgical and Materials Transactions A. 36A (2005) [111] J. Lyubina, O. Gutfleisch, R. Skomski, K.-H. Muller, L. Schultz, Phase transformations and thermodynamic properties of nanocrystalline FePt powders, i:10.1016/j.scriptamat.2005.04.041. [112] B. Zhang, T. K

lemmer, D. Hoydick, W.A. Soffa, the rela
lemmer, D. Hoydick, W.A. Soffa, the relationship between microstructure and magnetic properties in L1 ordered Fe-Pd ferromagnets, IEEE [113] A. Kulovits, J.M.K. Wiezorek, W.A. Soffa, W. Püschl, W. Pfeiler, The influence of deformation by cold-work on L1-ordering in anisotropic FePd, Journal of Alloys and Compounds. 378 (2004) 285–289. doi:10.1016/j.jallcom.2003.10.070. [114] T. Ichitsubo, M. Nakamoto, K. Tanakas, M. Koiwa, Effect of Applied Stress on Transformation of FePd Single Crystal, Materials Transactions. 39 (1998) [115] K. Tanaka, T. Ichitsubo, M. Koiwa, Effect of external fields on ordering of FePd, g: A. 312 (2001) 118–127. doi:10.1016/S0921-[116] J. Wiezorek, Cold-working and annealing of L1-ordering iron-palladium base intermetallics, Materials Science Forum. (2007) 1497–1494. [117] S. Farjami, T. Fukuda, T. Kakeshita, Effect of Magnetic Field on Microstructure Evolution during Disorder-Order Transformation in an Fe-Pd Alloy, Materials i:10.2320/matertrans.MAW200810. [118] D.R. Gaskell, Introduction to the thermodynamics of materials, Fifth Edition, LLC, New York, NY, 2008. [119] M.E. McHenry, D.E. Laughlin, Theory of magnetic phase transitions, in: E. Kaufmann (Ed.), Characterization of Materials, 2nd Editio, Wiley-Interscience, [120] P. Ehrenfest, Phase changes in the ordinary and extended sense classified according to the corresponding singularities of the thermodynamic potential., Proceedings

of the Koninklijke Nederlandse Akademie
of the Koninklijke Nederlandse Akademie van Wetenschappen. 36 [121] J.W. Gibbs, On the equilibrium of heterogeneous substances, Transactions of the Connecticut Academy of Arts and Sciences. III (n.d.). [122] W.A. Soffa, D.E. Laughlin, Recent experimental studies of continuous transformations in alloys: an overview, Proceedings of an International Conference on Solid-Solid Phase Transformations, AIME. (1982) 159–183. [123] D.E. Laughlin, K. Hono, Physical metallurgy: 3 - Volume set, 5t[124] J.B. Newkirk, A.H. Geisler, D.L. Martin, R. Smoluchowski, Ordering reaction in cobalt-platinum alloys., Jour[125] R.A. Ristau, K. Barmak, L.H. Lewis, K.R. Coffey, J.K. Howard, On the relationship of high coercivity and L1 ordered phase in CoPt and FePt thin films, [126] K. Barmak, B. Wang, A. Jesanis, D. Berry, J.M. Rickman, L1 FePt: Ordering, Anisotropy Constant and Their Relation to Film Composition, IEEE Transactions [127] K. Barmak, R.A. Ristau, K.R. Coffey, M.A. Parker, J.K. Howard, Grain growth and ordering kinetics in CoPt thin films, Journal of Applied Crystallography. 79 [128] Q.F. Xiao, E. Bruck, Z.D. Zhang, F.R. de Boer, K.H.J. Buschow, Phase transformation and magnetic properties of bulk CoPt alloy, Journal of Alloys and Compounds. 364 (2004) 64–71. do[129] J.M. Penisson, A. Bourret, P.. Eurin, Etude de la mise en ordre de l’alliage Pt-Co equiatomique par microscopie electronique, Acta Metallurgica. 19 (1971) 1195–

[130] C. Michaelsen, K. Barmak, T.P. Wei
[130] C. Michaelsen, K. Barmak, T.P. Weihs, Investigating the thermodynamics and kinetics of thin film reactions by differential scanning calorimetry, Journal of Physics D: Applied Physics. 30 (1997) 3167–3186. doi:10.1088/0022-[131] B. Wang, Ultrahigh density magnetic recording media: Quantitative kinetic experiments and models of the A1 to L1 phase transformation in FePt and related ternary alloy films, Carneg[132] J.M.D. Coey, Magnetism and magnetic materials, Cambridge University Press, [133] B.D. Cullity, C.D. Graham, Introduction to magnetic materials, 2nd Edition, IEEE/Wiley, NJ, 2009. [134] K.H.J. Buschow, F.R. de Boer, Physics of magnetism and magnetic materials, Kluwer Academic Publishers, 2003. [135] W. Pepperhoff, M. Acet, Constitution and Magnetism of Iron and Its Alloys, [136] D. Jiles, Introduction to magnetism and magnetic materials, 2nd Edition, Taylor [137] R. Skomski, Nanomagnetics, Journal of Physics: Condensed Matter. 841 (2003) R841–R896. [138] S.S.A. Razee, J.B. Staunton, B. Ginatempo, E. Bruno, F.J. Pinski, Ab initio theoretical description of the dependence of magnetocrystalline anisotropy on both compositional order and lattice distortion in transiti[139] S.S.A. Razee, J.B. Staunton, D.D. Johnson, B. Ginatempo, E. Bruno, Correlation of magnetocrystalline anisotropy of FePd alloy with chemical order, Journal of [140] S. Ostanin, S.S.A. Razee, J.B. Staunton, B. Ginatempo, E. Bruno, Magn

etocrystalline anisotropy and compositio
etocrystalline anisotropy and compositional order in Fe0.50.5: Calculations from an ab initio electronic model, Journal of Applied Crystallography. 93 (2003) [141] Y. Kota, A. Sakuma, Magnetocrysin FePt with L1 ordering and tetragonal distortion, Journal of Applied Physics. 111 (2012) 07A310. [142] C.J. Aas, L. Szunyogh, R.W. Chantrell, Effects of composition and chemical disorder on the magnetocrystalline anisotropy of Fe1-x alloys, Europhysics [143] S. Okamoto, N. Kikuchi, O. Kitakami, T. Miyazaki, Y. Shimada, K. Fukamichi, Chemical-order-dependent magnetic anisotropy and exchange stiffness constant of FePt(001) epitaxial films, Physical Review B. 66 (2002) 024413. [144] P. Kamp, A. Marty, B. Gilles, R. Hoffmann, S. Marchesini, M. Belakhovsky, et al., Correlation of spin and orbital anisotropies with chemical order in Fe0.50.5alloy films using magnetic circular x-ray dichroism, Physical Review B. 59 (1999) [145] K. Barmak, J. Kim, L.H. Lewis, K.R. Coffey, M.F. Toney, A.J. Kellock, et al., Stoichiometry–anisotropy connections in epitaxial L1 FePt(001) films, Journal of [146] S. Constantinides, The magnetic [147] H.L. Skriver, Materials Science Databases, Databases at CAMP. (2004). [148] E.J. Lavernia, T.S. Srivatsan, The raessing of materials: science, principles, technology, advances, and applications, Journal of Materials [149] C. Suryanarayana, Mechanical alloying and milling, Progress in Materials Science.

[150] J. Lyubina, O. Gutfleisch, K.-H.
[150] J. Lyubina, O. Gutfleisch, K.-H. Muller, L. Schultz, N.M. Dempsey, Nanocrystalline hard magnetic FePt powders, Journal of Applied Physics. 95 [151] B.Q. Han, J. Ye, A.P. Newbery, Y.T. Zhu, E.J. Schoenung, J. M. Lavernia, Bulk Nanostructured Metals from Ball Milling and Consolidation, in: M.J. Zehetbauer, Y.T. Zhu (Eds.), Bulk Nanostructured Materials, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2009. [152] B. Zhang, L. Lu, M.. Lai, Evolution of vacancy densities in powder particles during mechanical milling, Physica B: Condensed Matter. 325 (2003) 120–129. [153] K. Detemple, O. Kanert, J. De Hosson, K. Murty, In situ nuclear magnetic resonance investigation of deformation-generated vacancies in aluminum, Physical [154] T. Shintani, Y. Murata, Evaluation of the dislocation density and dislocation character in cold rolled Type 304 steel determined by profile analysis of X-ray diffraction, Acta Materialia. 59 (2011) 4314–4322. doi:10.1016/j.actamat.2011.03.055. [155] H. Alkhazraji, M.Z. Salih, Z. Zhong, M. Mhaede, H.-G. Brokmeier, L. Wagner, et al., Estimation of Dislocation Density in Cold-Rolled Commercially Pure Titanium tion, Metallurgical and Materials Transactions B. [156] J.I. Goldstein, D.E. Newbury, P. Echlin, D.C. Joy, A.D. Romig, C.E. Lyman, et al., opy and X-Ray Microanalysis, 2nd edition, Plenum [157] J.I. Goldstein, Scanning electron microscopy and x-ray microanalysis, KluwerAcademic/P

lenum, NY, 2003. [158] H.P. Klug, L.E. A
lenum, NY, 2003. [158] H.P. Klug, L.E. Alexander, X-ray diffraction procedures foamorphous materials, second edi, John Wiley & Sons, New York, 1974. [159] G.A. Novak, A.A. Colville, A practical interactive least-squares cell-parameter program using an electronic spreadsheet and a personal computer, American [160] I. Cernatescu, LPA and SAXS analysis, (n.d.). [161] G.K. Williamson, W.H. Hall, X-ray line broadening from filed aluminium and wolfram, Acta Metallurgica. 1 (1953) 22–31. [162] M.S. Senthil Saravanan, K. Sivaprasad, P. Susila, S.P. Kumaresh Babu, Anisotropy models in precise crystallite size determination of mechanically alloyed powders, Physica B: Condensed Matter. 406 (2011) 165–168. [163] H.M. Ledbetter, R.P. Reed, Elastic Propertand Iron-Nickel Alloys, Journal of Physical and Chemical Reference Data. 2 [164] A. Al-Ghaferi, P. Müllner, H. Heinrich, G. Kostorz, J.M.K. Wiezorek, Elastic constants of equiatomic L1-ordered FePd single crystals, Acta Materialia. 54 (2006) 881–889. doi:10.1016/j.actamat.2005.10.018. [165] B.D. Cullity, Elements of X-ray Diffraction, 1st ed., Addison-Wesley Publishing Company, Inc., Reading, MA, 1956. [166] B. Fultz, J. Howe, Diffraction and the X-Ray Powder Diffractometer, in: Transmission Electron Microscopy and Diffractometry of Materials, Springer [167] R. Pynn, Neutron scattering - A non-destructive microscope for seeing inside matter, in: L. Liyuan, R. Rinaldi, H. Sc

hober (Eds.), Neutron Applications in Ea
hober (Eds.), Neutron Applications in Earth, Energy and Environmental Sciences, Springer Science+Business Media, LLC, 2009. [168] S. V. Andreev, M.I. Bartashevich, V.I. Pushkarsky, V.N. Maltsev, L.A. Pamyatnykh, E.N. Tarasov, et al., Law of approach to saturation in highly anisotropic ferromagnets - Application to Nd-Fe-B melt-spun ribbons, Journal of Alloys and Compounds. 260 (1997) 196–200. [169] R. Grössinger, A critical examination of the law of approach to saturation, Physica [170] K. Barmak, K.R. Coffey, Metallic Films for Electronic, Optical and Magnetic Applications: Structure, Processing and Properties, Woodhead Publishing, Philadelphia, PA, 2014. [171] I. Beta, Theory of Thermogravimetry (TG) and Differential Scanning Calorimetry [172] M.S. Leu, C.S. Tsai, C.S. Lin, S.T. Lin, The determination of Curie temperature by differential scanning calorimetry under magnetic field, IEEE Transactions on [173] C. Takahashi, M. Ogura, H. Akai, First-principles calculation of the Curie temperature Slater-Pauling curve., Journal of Physics. Condensed MatterInstitute of Physics Journal. 19 (2007) 365233. doi:10.1088/0953-[174] Y. Kakehashi, Systematic Variations of Magnetic Properties in 3 Transition Metal Alloys, Progress of Theoretical Physics Supplement. (1990) 105–117. [175] E. V. Pal’guev, A.A. Kuranov, P.N. Syutkin, F.A. Sidorenko, Anomalous temperature dependence of the electrical resistance of ordered palladiu

m-iron [176] S. Ohba, Y. Saito, Y. Noda,
m-iron [176] S. Ohba, Y. Saito, Y. Noda, A measurement of charge asphericity in iron metal, [177] L.J. Swartzendruber, V.P. Itkin, C.B. Alcock, Fe-Ni (iron-nickel), in: H. Okamoto (Ed.), Phase Diagrams of Binary Iron Alloys, ASM International, Materials Park, [178] A. Nash, P. Nash, The Ni-Pd (Nickel-Palladium) system, Bulletin of Alloy Phase Diagrams. 5 (1984) 446–450. [179] P.R. Subramanian, D.E. Laughlin, Cu-Pd (Copper-Palladium), Journal of Phase. [180] M. Perzanowski, Y. Zabila, M. Krupinski, A. Zarzycki, A. Polit, M. Marszalek, Chemical order and crystallographic texture of FePd:Cu thin alloy films, Journal of [181] M. Krupinski, M. Perzanowski, a Zarzycki, Y. Zabila, M. MarszaFePdCu nanoisland arrays made by templated solid-state dewetting, [182] M. Fallot, Les alliages du fer avec les metaux de la famille du platine, Annales de [183] B. Sprusil, V. Sima, B. Chalupa, B. Smola, Phase transformations in CuAu and Cu3Au: a comparison between calorimetric and resistometric measurements, Zeitschrift Für Metallku[184] L. Battezzati, M. Belotti, V. Brunella, Calorimetry of orderi[185] J. Liu, K. Barmak, Interdiffusion in nanometric Fe/Ni multilayer films, Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films. 33 (2015) [186] K. Barmak, J. Kim, S. Shell, E.B. Svedberg, J.K. Howard, Calorimetric studies of the A1 to L1 transformation in FePt and CoPt thin films, Applied Physics Letters. [187] K. B

armak, J. Kim, D.C. Berry, W.N. Hanani,
armak, J. Kim, D.C. Berry, W.N. Hanani, K. Wierman, E.B. Svedberg, et al., Calorimetric studies of the A1 to L1 transformation in binary FePt thin films with compositions in the range of 47.5–54.4 at. % Fe, Journal of Applied Physics. 97 [188] J. Fillon, D. Calais, Autodiffusion dans les alliages concentres fer-palladium, J. Phys. Chem. Solids. 38 (1977) 81–89. [189] J.P. Gomez, C. Remy, D. Calais, Chemical diffusion and the Kirkendall effect in the system iron-palladium, Les Memoires Scientifiques de La Revue de [190] R.Z. Valiev, R.K. Islamgaliev, I. V. Alexandrov, Bulk nanostructured materials from severe plastic deforma[191] R. Valiev, Nanostructuring of metals by severe plastic deformation for advanced [192] J.M. Rickman, W.W. Tong, K. Barmak, Impact of heterogeneous boundary nucleation on transformation kinetics and microstructure, Acta Materialia. 45 [193] N.I. Vlasova, N.M. Kleinerman, V. V. Serikov, A.G. Popov, Mössbauer study of fine structure features of equiatomic FePd alloy after severe plastic deformation and ordering annealing, Journal of Alloys and Compounds. 583 (2014) 191–197. doi:10.1016/j.jallcom.2013.08.110. [194] H. Okamoto, D.J. Chakrabarti, D.E. Laughlin, T.B. Massalski, The Au-Cu (Gold-Copper) System, Bulletin of Alloy Phase Diagrams. 8 (1987) 454–473. [195] D.B. Butrymowicz, J.R. Manning, M.E. Read, Diffusion in Copper and Copper Alloys, Part II. Copper-Silver and Copper-Gold Systems,

J. Phys. Chem. Ref. [196] Y. Feutelais,
J. Phys. Chem. Ref. [196] Y. Feutelais, B. Legendre, M. Guymont, New enthalpies determination in situ X-ray diffraction observations of order/disorder transitions in Au0.50.5, Acta [197] B.Q. Zhang, L. Lu, M.O. Lai, Evolution of vacancy densities in powder particles during mechanical milling, Physica B: Condensed Matter. 325 (2003) 120–129. [198] M.J. Zehetbauer, G. Steiner, E. Schafler, A. Korznikov, E. Korznikova, Deformation induced vacancies with severe plastic deformation: measurements and modeling, Materials Science Forum. 503-504 (2006) 57–62. [199] D. Setman, E. Schafler, E. Korznikovaof vacancy type defects in nanometals detained by severe plastic deformation, neering A. 493 (2008) 116–122. doi:10.1016/j.msea.2007.06.093. [200] E. Schafler, G. Steiner, E. Korznikovainvestigation of ECAP-Cu by means of X-ray line profile analysis, calorimetry and electrical resistometry, Materials Science and Engineering A. 411 (2005) 169–173. doi:10.1016/j.msea.2005.08.070. [201] A.C. Larsen, R.B. Von Dreele, General Structure Analysis System (GSAS). Los Alamos National Laboratory Report LAUR 86-748, Los Alamos, New Mexico, [202] B.H. Toby, EXPGUI , a graphical user interface for GSAS, Journal of Applied [203] S. Lefebvre, F. Bley, M. Fayard, M. Roth, Neutron diffuse scattering investigation of different states of local order in 0.7650.235, Acta Metallurgica. 29 (1981) [204] J. Orehotsky, Magnetic and crystallographic c

haracteristics of the order-disorder pha
haracteristics of the order-disorder phase transformation in Pt, IEEE Transactions on Magnetics. 25 (1989) 3791–[205] J. Orehotsky, J.L. Orehotsky, Kinetics of the order-disorder transformation in [206] J. Orehotsky, J.L. Orehotsky, An anomalous magnetic effect in the Pt order–disorder alloy system, Journal of Applied Physics. 61 (1987) 4240. [207] N. Bordeaux, Structure-Magnetism Correlations and Chemical Order-Disorder Transformations in L1-Structured Compounds, Northeastern University, 2015. [208] L. Kraus, J. Bydžovský, P. Švec, Continuous stress annealing of amorphous ribbons for strain sensing Actuators A. 106 (2003) 117–[209] S.J. Kernion, P. Ohodnicki, J. Grossmann, A. Leary, S. Shen, V. Keylin, et al., Giant induced magnetic anisotropy in strain annealed Co-based nanocomposite [210] G. Herzer, V. Budinsky, C. Polak, Magnetic properties of nanocrystalline FeCuNbSiB with huge creep induced anisotropy, Journal of Physics:Conference [211] Burton, Amorphous metallic alloys. Butterworths monographs in materials, [212] R. Hasegawa, K. Takahashi, B. Francoeur, P. Couture, Magnetization kinetics in tension and field annealed Fe-based amorphous alloys, Journal of Applied Physics. [213] M.E. McHenry, Nanocomposite alloy design for power elect[214] M.A. Willard, J.H. Claassen, V.G. Hams, Magnetic and Structural Properties of Amorphous and Nanocrystalline FeNi-based Alloys, in: Proceedings of the 2001 2001. IEEE-N

ANO 201, 2001: pp. 51–[215] C. Chang, B.
ANO 201, 2001: pp. 51–[215] C. Chang, B. Shen, A. Inoue, FeNi-based bulk glassy alloys with superhigh mechanical strength and excellent soft- magnetic properties, Applied Physics [216] R.W. Cahn, Lattice parameter changes on disordering intermetallics, Intermetallics. 7 (1999) 1089–1094. [217] N. Braidy, Y.L. Bouar, M. Fèvre, C. Ricolleau, Determination of the long-range order parameter from the tetragonality ratio of L1 alloys, Physical Review B. 81 [218] B.W. Roberts, X-ray measurement of order in CuAu, Acta Metallurgica. 2 (1954) [219] V.D. Sukhanov, O.D. Shashkov, T.S. Boyarshinova, Y.E. Turkhan, Atomic ordering and solubility in substitutional solid solution: I. The A1-L1 ordering, The [220] D.E. Cox, High-resolution powder diffraction and structure determination, in: P. Coppens (Ed.), Synchrotron Radiation Crystallography, Academic Press, San [221] P. Stephens, Principles and applic[222] Lawrence Berkeley National Laboratory, ed., X-ray data BUniversity of California, Berkeley, CA, 2001. [223] J.-L. Hodeau, V. Favre-Nicolin, S. Bos, H. Renevier, E. Lorenzo, J.-F. Berar, Resonant Diffraction, Chemi[224] B. Wang, K. Barmak, Re-evaluation of the impact of ternary additions of Ni and Cu on the A1 to L1 transformation in FePt films, Journal of Applied Physics. 109 [225] A. Djekoun, N. Boudinar, A. Chebli, A. Otmani, M. Benabdeslem, B. Bouzabata, et al., Structure and magnetic properties of Fe-rich nanostructur

ed Fe100-Xpowders obtained by mechanical
ed Fe100-Xpowders obtained by mechanical alloying, Physics Procedia. 2 (2009) 693–700. [226] K. Gheisari, J.T. Oh, S. Javadpour, The effect of heat treatment on the structure and magnetic properties of mechanically alloyed Fe–45%Ni nanostructured powders, Journal of Alloys and Compounds. 509 (2011) 1020–1024. doi:10.1016/j.jallcom.2010.09.159. [227] A. Guittoum, A. Layadi, A. Bourzami, H. Tafat, N. Souami, S. Boutarfaia, et al., X-ray diffraction, microstructure, Mössbauer and magnetization studies of alloy prepared by mechanical alloying, Journal of Magnetism and Magnetic Materials. 320 (2008) 1385–1392. doi:10.1016/j.jmmm.2007.11.021. [228] E. Jartych, J.K. urawicz, D. Oleszak, M. Pa, X-ray diffraction, magnetization and Mössbauer studies of nanocrystalline Fe–Ni alloys prepared by low- and high-energy ball milling, Journal of Magnetism and Magnetic Materials. 208 (2000) 221–230. doi:10.1016/S0304-8853(99)00543-0. APPENDICES This Section contains background information that complements calculations and analyses performed in this dissertation. Appendix 1 presents the method used to account for the instrumental contribution to the breadth of X-ray diffraction peaks. Appendix 2 describes the factors involved in the calculation of the theoretical integrated intensity of x 3 explains the method used to correct the intensity of the X-ray diffraction peaks to account for finite-sample-size effects. Appendix 4 pres

ents the results of a study performed to
ents the results of a study performed to optimize the cryomilling procedure of FeNi-based alloys to attain high strain levels. Appendix 1 Instrumental correction for X-ray peak broadening analysisThe instrumental contribution to the Bragg XRD peak breadth must be accounted for when performing peak breadth analysis via the Williamson-Hall method or any other related technique on X-ray data obtained as a function of . In this manner, a more accurate determination of crystallite size and microstrain are obtained. The instrumental peak breadth contribution is determined from the measurement of a known polycrystalline untextured powder standard with an average crystallite size much larger than 100 nm and with negligible microstrain, measured under identical conditions as the samples of interest. The peak breadth of each individual Bragg reflection of the standard ), obtained as the full-width at half-maximum (FWHM), is used to generate a plot of A linear regression to the dataprovides the instrumental contribution to peak breadth ( at any angle. In this dissertation, the instrumental contribution to peak breadth for the laboratory X-ray experiments was obtained by measuring a silicon In synchrotron X-ray diffraction experiments, if a powder standard is not available, an approximation to the variation of instrumental breadth with scattering angle can be obtained from the angular resolution ation with a crystal (

35 ) Where: of the incident beam, typic
35 ) Where: of the incident beam, typically 0.005° (deg) = contribution of the Darwin widths (intrinsic width) of the monochromator and crystal analyzer, calculated as the square root of the sum of the squared values (deg) At NSLS (BNL) beamline X-16C, employed in this dissertation for the characterization of FeNi-based materials (described in Section 4.3.3.1), the experimental setup included a Si() monochromator and a Ge() analyzer, as shown in Figure 84. The corresponding Bragg angles and Darwin widths for these crystals are reported in Table 11. The instrumental breadth determined for data obtained from the synchrotron X-ray experiments, calculated employing Equation 35, is small but not negligible, varying Figure 84. Instrumental setup at the X-16C beamline at NSLS (BNL). Image age Table 11. Bragg angles and Darwin widths of Si( (degrees) 14.2206 13.6416 Darwin width (degrees) 0.00188 0.00455 Once an estimate of the instrumental breadth contribution at a particular scattering angle is obtained, it can be used to correct the measured peak breadth () for the Bragg reflections of the sample, and in this way obtain the sample’s contribution to peak breadth ( ( 36 ) Where: = sample contribution to peak breadth (deg) = measured breadth (full-width at half-maximum) for the sample (deg) Appendix 2 Determination of the theoretical integrated intensity of X-ray diffraction peaks in a sample is necessary to

calculate the degree of long-range orde
calculate the degree of long-range order in L1 chemically ordered systems or the phase fraction in multi-phase systems. In general terms, the theoretical integrated X-ray diffraction intensity () of a Bragg peak is given by [81]: ( 37 ) Where: charge and mass of electron c = speed of light = radius of the diffractometer = volume of unit cell = linear absorption coefficient = structure factor = multiplicity factor for the Bragg reflection under consideration = Polarization factor = scattering angle An explanation of each of the factors included in the expression for together with a method for its calculation, will be given next. The structure factor is a mathematical description of how a crystal scatters incident radiation. by adding all the waves scattered by individual atoms (1, 2, 3, …, N) in the unit cell with ) and atomic scattering factor ¦NlwkvhuinnnnefF12 ( 38 ) is a complex number that expresses both the amplitude and phase of the resultant wave. The intensity of the beam diffracted by all atoms of the unit cell in a direction predicted by Bragg’s law is proportional to the magnitude of the structure factor ||. The multiplicity factor takes into account the relative proportion of planes contributing to the same reflection ( with the same -spacing) [81]. For instance, in a polycrystalline sample with a

cubic lattice, the () planes will diffra
cubic lattice, the () planes will diffract at the same angle as the (), (), (), (), and (for the {} plane family. The intensity of this Bragg peak is proportional to the number of planes contributing to it, and thus a mushould be applied. accounts for the length of time that a crystal remains in the diffracting position for any particular scattered beam. This depends on both the Bragg angle and the diffraction geometry. The integrated peak intensity of the diffracted beam is ) [158,165]. The polarization factor XRD stems from the fact that unpolarized radiation traveling in the -direction towards a crystal at the proper angle for diffraction becomes polarized after being scattered, with the amount of polarizscattering angle [158]. It originates from the difference in angles between the and components of the electric vector of the incident radiation and the direction of scattering, and is given by XRD)/2. A combined Lorentz-Polarization factor (commonly used for intensity calculations, with the constant 1/8 omitted [81]: ( 39 ) Absorption of X-rays in the specimen also affects the intensity of X-ray diffraction peaks. For a flat plate configuration, common in laboratory diffractometers, X-ray absorption is independent of the diffraction angle, provided the sample covers the entirety of the X-ray illuminated area and is effectively of infinite thickness. The is the linear absorptitiLastly, the temperatu

re factor accounts for thermal vibration
re factor accounts for thermal vibrations of atoms in the crystal lattice, which smear out the lattice planes and thus result in an overall decrease in the intensity of the diffracted beam. The temperature factor is expressed as exp[- is the isotropic temperature factor. For nd is typically taken as 1 [158]. When comparing the intensity of two different Bragg peaks in a diffraction pattern corresponding to the same phase, for example for long-range order calculation, there are several factors that apply equally to both Bragg reflections and thus do not need to be considered in the calculation of . In such cases, the theoretical intensity equation can be simplified as: ( 40 ) If on the other hand a comparison of integrated intensities is being made between Bragg peaks in a diffractogram corresponding to different phases, for example for phase fraction calculation, the theoretical intensity equation is simplified as follows: ( 41 ) To finalize this Appendix, here are some considerations that were taken in this lation of theoretical integrated intensities: The calculation of the structure factor for fundamental and superlattice reflections in an L1 lattice considers the fact that in this crystal structure atom A occupies the lattice positions (0, 0, 0) and (½, ½, 0) in the lattice, while atom B occupies positions (½, 0, ½) and (0, ½, ½). Thus, th ( 42 ) BAfundffF unmixed (fundamental lines) ( 43 ) BAffF m

ixed (superstructure lines) ( 44 ) For a
ixed (superstructure lines) ( 44 ) For an AB alloy with a partially ordered L1 lattice, considering X and X as the atomic fractions of A and B, respectively, ( 45 ) BAffLROF ( 46 ) Where is the long-range order parameter. The calculation of the structure factor for a bcc lattice considers the fact that in this crystal structure the atoms occupy lattice ( 47 ) ( 48 ) even) ( 49 ) It was seen above that the intensity of Bragg peaks depends on the structure factor which is a function of the scattering factor of the atoms in the material. The atomic scattering factor of any element is usually considered to be independent of the of the incident radiation provided the quantity (sin is constant [81]. However, when is near the wavelength for the -absorption-edge of the scattering element (), the atomic scattering factor changes by an amount , resulting in an overall which may be several units lower than it is when en . Δf is the resonant (or anomalous) scattering factor which depends on the chemical nature of the scattering element [223]. Considering a binary AB alloy that undergoes the A1 phase transformation, (where A and B are the constituent elements), it was seen in Section 2.1.3.2 that superstructure peaks characteristic of chemical order appear in upon formation of the L1 phase. The intensity of the L1 superstructure peaks, as demonstrated in this Appendix, depends on the difference in scatteri

ng factors of the A and B elements. For
ng factors of the A and B elements. For an L1-structured FeNi alloy the similarity in scattering factors of Fe and Ni results in superlattice peaks that are undetectable above the noise level (Section 2.1.3.5). However, if the wavelength of the diffraction experiment is tuned to the absorption edge of Fe or Ni under so-called anomalous diffrscattering factor of that element may be significantly varied to donate greater Fe/Ni the intensity of the superlattice peaks can be enhanced. Appendix 3 XRD Correction factor for finite size sample Appendix 2 described how the intensity of a diffracted X-ray beam is attenuated by absorption effects, which are independent of scattering angle in a flat plate configuration provided the sample covers the entirety of the X-ray illuminated area and is effectively of infinite thickness . The infinite thickness criterion is given as follows [81]: ( 50 ) If the sample can be considered to be infinitely thick, but does not cover the entirety of the X-ray illuminated area, then the volume fraction of the specimen that is contributing to the scattering of X-rays varies with angle, resulting in a variation of the absorption factor. In such cases, a correction of the experimentally obtained integrated intensities must be made to account for the finite sample size. This correction can be accomplished by measuring two polycrystalline, untextured standard materials in XRD, one of them wi

th an area that is sufficient to cover t
th an area that is sufficient to cover the X-ray beam (“large” standard), and another of area similar to the sample of interest (“small” standard). The measurement must be performed under identical conditions as those used for the sampa plot of the ratio of the integrated intensities of large and small standards (large will provide a means to estimate an area correction factor for each Bragg reflection of the sample [224]. In this dissertation, the correction factor was determined by comparison of the experimental integrated intensities of the Bragg reflections of two polycrystalline randomly oriented Si powder disk references, with diameters ~3 mm (equal to diameter of the FePdM samples of Aim 1) and ~22 mm (larger than X-ray beam illuminated area). These references were fabricated from Si powder (AlfaAesar) mixed with collodion. Care was taken to ensure that the thickness of both Si-large and Si-small standards was in accordance with the infinite thickness criterion. Figure 85a and b show the XRD patterns for Si-small and Si-large standards. All intensities were normalized to the intensity of the () peak. The ratio (was fit to an exponential function (Figure which was applied to the experimentally obtained integrated intensities of the Ø 3 mm FePdM disk samples. Figure 85. X-ray diffractograms of polycrystalline untextured powder Si standards for the calculation of the area correction factor. a) Small samp

le, with a size similar to samples of in
le, with a size similar to samples of interest. b) Large sample, with an area sufficient to cover completely the incident X-ray beam. 228 Figure 86. Plot of largesmall vs. for the Si standard. Data points were fit to an exponential function to determine the correction factor that should be applied to the integrated intensity of Bragg reflections at any given for Appendix 4 Optimization of cryomilling procedure for FeNi-based alloys Prior to the examination of the potential of cryomilling techniques to favor the formation of the L1 phase in FeNi-based alloys, the cryomilling process was optimized to achieve the highest microstrain levels in reasonable laboratory times. FeNi melt-spun ribbons were cryomilled following the same procedure described in Section 4.3.2.1, which in brief consists of mixing the cut ribbons with surfactants and using milling cycles composed of 10 min of milling at a rate of 15 cycles/second followed by 2 min of cooling. At cumulative milling times of 1, 5, and 9 h part of the sample was extracted, cleaned to remove surfactants, and characterized by X-ray diffraction. The diffractograms obtained for the cryomilled samples and the Within the limits of detection, all samples were found to be in a single fcc-phase. With increasing milling time, the Bragg peaks experienced broadening, which was monitored and analyzed through a modified Williamson-Hall technique (described in Section 3.3.

2). After correction for instrumental br
2). After correction for instrumental breadth contribution (see Appendix 1 for vs reveals that the as-spun sample is independent, indicating that the peak broadening observed is attributed to crystallite size only, Figure 88. As milling proceeds, the data becomes dependent, with both crystallite size and microstrain contributing to peak breadth. From the slopes and intercepts of the linear fits to the data, estimates of the crystallite size and microstrain where obtained, and the illing time was determined, Figure 89. Figure 88. Modified Williamson-Hall plot for FeNi cryomilled for different times. The broken lines represent the linear fit to the data, from which the intercept and slope are extracted to estimate the crystallite size and microstrain. Cryomilling for 1 h produces a drastic decrease in crystallite size and increase in microstrain, after which additional cryomilling produces modest changes in these parameters. Even though the incremental change in microstrain per additional milling hour is low after the first hour of cryomilling, it was determined that a milling time of 9 h (1-work day) was reasonable in terms of processing time spent, and allowed the achievement of high microstrain levels at values of maximum strain levels obtained by traditional dry ball-milling of FeNi powder mixtures include 0.185% for Fe milled for 97 h [225], 0.26% for Fe milled for 24 h [226], 0.41% for Fe milled for 5