Synthesis and Characterization of Superconducting Ferropnictide Bulks and Wires Jeremy Weiss

Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2015 Synthesis and Characterization of Superconducting Ferropnictide Bulks and Wires Jeremy Weiss

Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected] FLORIDA STATE UNIVERSITY

THE GRADUATE SCHOOL

SYNTHESIS AND CHARACTERIZATION OF SUPERCONDUCTING FERROPNICTIDE

BULKS AND WIRES

JEREMY WEISS

A Dissertation submitted to the Department of Materials Science and Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Degree Awarded: Spring Semester, 2015 Jeremy Weiss defended this dissertation on April 9, 2015. The members of the supervisory committee were:

Eric Hellstrom Professor Directing Dissertation

Gregory Boebinger University Representative

David Larbalestier Committee Member

Theo Siegrist Committee Member

Per Arne Rikvold Committee Member

The Graduate School has verified and approved the above-named committee members, and certifies that the dissertation has been approved in accordance with university requirements.

This dissertation is dedicated to the memory of Professor James Brooks who served on this dissertation advisory committee until his passing in Fall 2014

iii

ACKNOWLEDGMENTS

Firstly, I would like to thank David Larbalestier for giving me the undergraduate opportunity to do the preliminary lab work that sparked an interest in applied research and ultimately lead to the following dissertation. To vaguely paraphrase the first toast of his I heard; an observation relevant to my developing career: “We never could have guessed that we would have ended up here, but we stumble through life as opportunities present themselves and it really is fascinating how it seems to work out for us.” It was David who introduced me to Jianyi Jiang who mentored me for two and a half years as an undergraduate. Jianyi showed an unfounded amount of patience as I learned to succeed (and fail) at research, was always willing to answer any question I brought him, taught me almost everything I know about electromagnetic characterization, and gave me the freedom to explore new techniques and procedures. I owe a lot of gratitude to Eric Hellstrom who mentored me as a graduate student. Eric taught me the most important skills I now possess including the ability to communicate effectively. With his help I have come from being a mess when presenting to winning multiple best presentation awards. I want to thank Bill Starch for purchasing, fixing, training, and managing of all things technical. When I first came to the ASC, I coveted Bill’s technically demanding job, and within a couple of years he trusted me enough to attempt almost any task I dared to take on, but there is only one Bill Starch. I’d like to thank the many scientists at ASC that have provided support. Chiara Tarantini provided many thoughtful discussions about fundamental theory and experimental techniques. Anitolii Polanskii, Fumitake Kametani, Dmytro (Dima) Abraimov, Van Griffin, and Akiyasu Yamamoto shared their characterization expertise and support. Ashleigh Francis, Muriel Hannion, Ben Hainsey, Ross Richardson, Matthieu Dalban-Canassy, Julian Velasquez, Jeff Whalen, Tiglet Basara, and Michael Santos all provided much needed technical support. Connie Linville and Charlotte Hall both went above and beyond when it came to providing wonderful administrative support. Jorge González, José Moreno, Marcos Corchado, and Gerardo Nazario were all National Science Foundation (NSF) sponsored research experience for undergraduate students whom I mentored over the summers on unique projects.

I’d like to acknowledge the special opportunities I was awarded over the years, enabled by Eric Hellstrom and David Larbalestier, including financial support to attend summer schools funded by the Institute for Complex Adaptive Matter (ICAM), the National High Magnetic Field

(NHMFL) Laboratory, and the FSU Research Foundation. I would also like to acknowledge supplemental financial support to attend conferences from the American Ceramic Society, and the NHMFL. I am grateful to the Institute of Electrical and Electronics Engineers (IEEE) Council on Superconductivity for awarding me a fellowship, without which ensuing unmanageable credit card debt was sure to become an extra burden. Last, and certainly not least, this work was supported by NSF DMR-1306785, NSF DMR-1006584, by the NHMFL which is supported by the NSF under NSF/DMR-0084173 and NSF DMR-1157490, and by the State of Florida. Work at the Atominstitut has been supported by the Austrian Science Fund (FWF): 22837-N20 and the European-Japanese collaborative project SuperIron (No. 283204). Atom- probe tomography measurements were performed at the Northwestern University Center for Atom-Probe Tomography (NUCAPT) and the LEAP tomograph was purchased and upgraded with funding from the NSF-MRI (DMR 0420532) and ONR-DURIP (N00014-0400798, N00014-0610539, N00014-0910781) programs. NUCAPT is a Shared Facility of the Materials Research Center of Northwestern University, supported by the National Science Foundation's MRSEC program (DMR-1121262). We are also grateful to the Initiative for Sustainability and Energy at Northwestern for upgrading NUCAPT’s capabilities. The work at the University of Tokyo was supported by the Japan Science and Technology Agency, PRESTO.

I’d also like to individually acknowledge the members of my supervisory committee for their input, expertise, and guidance. Theo Siegrist taught me much about crystallography and a bit about interfacing with machines older than myself. Steven Van Sciver taught me everything I know about cryogenic and magnet design, and it was a great pity he was not available to see this dissertation through to the end. I’d like to thank Per Arne Rikvold for happily agreeing to serve on this committee. James Brooks was an exemplary role model for what every scientist and educator should be. He took on more work than anybody should while maintaining his trademark sense of humor about it. It was a tragedy to see him go, but I am thankful to have known him and am fortunate Greg Boebinger was willing to take his place on this committee.

TABLE OF CONTENTS

List of Tables ...... x List of Figures ...... xi List of Symbols, Acronyms, and their Meanings ...... xvii Abstract ...... xx 1 - Introduction ...... 1 1.1 – Background ...... 2

1.2 – Parallels Between FBS and Cuprates ...... 5

1.3 – Current Transport in Bulk Ferropnictides ...... 6

1.4 – Characteristic Lengths in Type-II Superconductors ...... 8

1.5 – Bean’s Critical State Model ...... 10

2 - Experimental Techniques ...... 11 2.1 – Introduction ...... 11

2.2 – Safety Considerations ...... 11

2.3 – Synthesis Techniques ...... 12

2.3.1 – Ambient pressure solid-state synthesis ...... 12

2.3.2 – Mechanochemical synthesis...... 13

2.3.3 – Hot isostatic pressing ...... 13

2.3.4 – Wire fabrication ...... 14

2.3.5 – Large bulk fabrication ...... 14

2.4 – Electromagnetic Characterization ...... 16

2.4.1 – SQUID magnetometry ...... 16

2.4.2 – M vs T measurements ...... 16

2.4.3 – Trapped remanent field measurements ...... 17

2.4.4 – VSM magnetometry ...... 19

2.4.5 – M vs. H measurements ...... 20

2.4.6 – Physical property measurements...... 21

2.4.7 – R vs. (T, H) measurements ...... 21

2.4.8 – Ic vs. H measurements ...... 22

2.4.9 – Magneto optical imaging (MOI) ...... 23

2.5 – Microstructural Characterization ...... 23

2.5.1 – Optical imaging ...... 23

2.5.2 – Scanning electron microscopy (SEM) ...... 24

2.5.3 – X-ray diffractometry (XRD) ...... 24

2.5.4 – Energy-dispersive x-ray spectroscopy (EDS) ...... 25

2.6 – Nanostructural Characterization ...... 25

2.6.1 – Transmission electron microscopy (TEM) ...... 25

2.6.2 – Atom probe tomography (APT) ...... 25

3 - Mechanochemical Synthesis of Pnictide Compounds ...... 27 3.1 – Introduction ...... 27

3.2 – Experimental Details ...... 29

3.3 – Results ...... 31

3.4 – Discussion ...... 35

3.5 – Conclusions ...... 37

4 - High Intergranular Current Density in Fine Grain Ferropnictides ...... 38 4.1 – Introduction ...... 38

4.2 – Experimental Details ...... 39

4.3 – Results ...... 41

4.4 – Discussion ...... 46

5 - Dependence of K-doped BaFe2As2 Superconducting Properties on Sintering Temperature in Wires and Tapes ...... 49 5.1 – Introduction ...... 49

vii

5.2 – Experimental Details ...... 50

5.3 – Results ...... 50

5.4 – Discussion ...... 55

5.5 – Conclusions ...... 56

6 - Understanding Weak Links in Co and K-doped BaFe2As2 ...... 57 6.1 – Introduction ...... 57

6.2 – Experimental Details ...... 58

6.3 – Results and Discussion ...... 59

6.3.1 – The chemistry of grain boundaries ...... 59

6.3.2 –Intergranular vs intragranular magnetization ...... 60

6.3.3 – Percolation of current paths ...... 63

6.3.4 – Irreversible intergranular critical current density in applied fields ...... 64

6.4 – Conclusions ...... 66

7 - Evidence for Composition Variation and Impurity Segregation at Grain Boundaries in High Current Density Polycrystalline K- and Co-doped BaFe2As2 Superconductors ...... 67 7.1 – Introduction ...... 68

7.2 – Experimental Details ...... 69

7.3 – Results and Discussion ...... 70

7.4 – Conclusions ...... 76

8 - Demonstration of an Iron-Pnictide Bulk Superconducting Magnet Trapping Over 1T ...... 77 8.1 – Introduction ...... 77

8.2 – Experimental Details ...... 78

8.3 – Results ...... 79

8.4 – Discussion ...... 83

8.5 – Conclusions ...... 86

9 - Exploration of Paths Toward Higher Critical Current Densities in K-doped BaFe2As2 Polycrystals ...... 87

viii

9.1 – Introduction ...... 87

9.2 – Experimental Details ...... 88

9.3 – Results ...... 89

9.4 – Discussion ...... 92

9.5 – Conclusions ...... 94

Appendix A - Processing Iron Based Superconductors Safely ...... 95 Appendix B - Integrated Safety Management Plan ...... 98 Appendix C - Production of Multifilamentry K-doped Wires ...... 100 Appendix D - Copyright Permission Letter ...... 103 References ...... 105 Biographical Sketch ...... 119

LIST OF TABLES

Table 3.1 - Material properties of superconducting bulks...... 32

Table 7.1 - Nominal and APT measured compositions of bulk Ba122 samples (at.%) ...... 71

Table 7.2 - Nominal and APT measured compositions inside crystallites (grains) within Ba122 samples (at.%) ...... 71

Table 8.1- Comparison of superconducting and mechanical properties for YBCO, MgB2 and K-doped Ba122 ...... 85

Table 9.1 – Superconducting properties of samples measured ...... 92

LIST OF FIGURES

BGB Figure 1.1 - Dependence of the critical current density across GBs Jc as a function of 14 16 the [001] tilt misorientation angle for Co-doped BaFe2As2, P-doped BaFe2As2, and YBCO.17 ...... 4

Figure 1.2 - Upper critical field as a function of temperature for various superconductors.18 ...... 4

Figure 1.3 – Optical image taken under polarized light (left) of a polycrystalline Ba(Fe0.9Co0.1)2As2 bulk sample and the corresponding magneto optical image (right) taken after Zero-field cooling and applying a magnetic field of 100 mT at 7 K...... 6

Figure 2.1 – Remanent magnetic moment as a function of maximum applied field for two Bi-2223 flat wires before bending and after bending to suppress the contribution of local local mR . Dashed line represents bent wire mR subtracted from the unbent wire mR. Schematic in top left corner shows the hysteretic nature of self-field Jc as a function of maximum applied field. Data taken from ref.54 ...... 19

Figure 2.2 - Schematic of atomic probe microscopy setup. Arrows indicate the direction of evaporated ions...... 26

Figure 3.1 - SEM image of BaFe2As2 made by hand grinding Fe, Ba3As2, and As followed by HIP treatment at 1120 °C. The image shows voids and Fe2As that wets grain boundaries...... 28

Figure 3.2 - Milling vial temperature as a function of milling time for three different MSR reactions. The baseline was obtained by milling fully-reacted BaFe2As2 powder...... 30

Figure 3.3 - (a) XRD pattern of milled 0.6 Ba + 0.4 K + 2 Fe + 2 As powder before MSR. XRD pattern and SEM image of (b) milled powder after MSR showing that Ba0.6K0.4Fe2As2 had formed, (c) after the 1120 °C HIP heat treatment of MSR powder, (d) after the 600 °C AP heat treatment of MSR powder, and (e) after the second 600 °C HIP heat treatment of the MSR powder...... 31

Figure 3.4 - Temperature dependence of the volumetric susceptibility under zero field cooling (ZFC) in an external field of 20 Oe for (Ba0.6K0.4)Fe2As2 samples 1120 ° C HIP, 600 ° C AP, and 600 ° C HIP...... 32

Figure 3.5 - Magneto optical images showing flux penetration after zero-field-cooling (ZFC) the sample to 10 K and applying the magnetic field shown in the images, for samples (a) 1120 ° C HIP, (c) 600 ° C AP, and (e) 600 ° C HIP. (b), (d), and (f) are magneto-optical images of the remanent magnetic flux after the field was removed. These correspond to (a), (c), and (e), respectively...... 33

Figure 3.6 - Magnetic field dependence of the critical current density calculated from magnetization measurements at 4.2, 10, 15, 20, 25, and 30 K for samples (a) 600 ° C AP, and (b) 600 ° C HIP...... 34

Figure 4.1 - Volumetric magnetic susceptibility as a function of temperature for K-doped Ba122 wire and bulk. The magnetic response was evaluated by warming above Tc after zero field cooling to 5 K and applying a field of 2 mT parallel to the sample’s length...... 40

Figure 4.2 - Resistivity measurements of K-doped Ba122 bulk material. (a) Temperature dependence of resistance at different magnetic fields up to 35 T. The trend of resistivity with respect to applied field is very similar to that of K-doped single crystals , even though our bulk is untextured. It has ρ(300 K & 39 K) = (0.48 mΩcm & 0.07 mΩcm) compared to ρ(300 K & 39 K) = (0.6 mΩcm & 0.05-0.12 mΩcm) for single crystals REF, indicating that the normal-state properties are not being degraded by the presence of grain boundaries. Inset is 0 T resistivity up to 300K and RRR is ρ(300 K) divided by ρ(39 K).

(b) Hc2(T) defined at 90% (H90), 50% (H50) and 10% (H10) resistance...... 41

Figure 4.3 - Upper critical field as a function of temperature. (a) Hc2(T) defined at 90% resistance for the K-doped Ba122 bulk compared to an optimally doped single crystal 2 22 from reference, a Nb3Sn wire from reference, and a textured MgB2 thin film from reference81 with H applied parallel (closed symbols) and orthogonal (open symbols) to its 2 surface. The dotted line is a rescaled fit from reference to guide the eye. (b) Hc2 and temperature normalized by Tc to show close agreement between bulk polycrystal and single crystal with H//ab...... 42

Figure 4.4 - Microstructures of K-doped Ba122 bulk investigated by TEM. (a) TEM image of polycrystalline bulk K-doped Ba122 material showing several equiaxed grains with average grain diameter of ~200 nm. Inset is a selected area electron diffraction image of a that indicates that the grains of the material are randomly oriented with many high-angle grain boundaries. (b) HRTEM image of a typical K-doped Ba122 grain boundary where the TEM sample was tilted so the electron beam was almost parallel to the GB plane. The lattice fringes of upper and bottom grains meet at the grain boundary without an amorphous contrast, indicating the grain boundary is clean without a wetting impurity phase...... 43

xii

Figure 4.5 - Powder X-ray diffraction of K-doped Ba-122 bulk material. Small FeAs peaks can be seen from the impurity phase that occupies less than 3% of the sample volume without significantly blocking current...... 43

Figure 4.6 - Microstructure of K-doped Ba-122 bulk and wire investigated by SEM. (a) SEM image of polycrystalline bulk K-doped Ba-122 material showing non grain-wetting FeAs impurity phase. (b) SEM image of the K-doped Ba-122 wire’s superconducting cross section showing non grain-wetting FeAs impurity phase. FeAs phase accounts for less than 3% of the cross sectional area by image analysis in both bulk and wire...... 44

Figure 4.7 - Microstructure of Co-doped Ba122 wire investigated by TEM. TEM image of polycrystalline bulk Co-doped Ba122 material showing equiaxed grains with average grain diameter less than 200 nm. Inset is a selected area electron diffraction image that indicates the grains of the material are randomly oriented with many high-angle grain boundaries. TEM confirms the Co-doped wire is structurally comparable to the K-doped wire with many well connected grains...... 44

Figure 4.8 - Optical and Magneto-optical images of a K-doped Ba122 wire cross section with magnetic fields applied perpendicular to the shown cross section. (a) Optical image of the wire cross section showing superconducting core surrounded by Ag and Cu sheath. (b) Magneto-optical image of trapped magnetic flux in the wire field-cooled (FC) to 7 K in an external magnetic field of 120 mT...... 45

Figure 4.9 - Magneto-optical images of a rectangular piece of K-doped Ba122 bulk material with magnetic fields applied perpendicular to plain of the sample (thickness = 0.7 mm). (a) Magneto-optical image of partial flux penetration after zero-field-cooling (ZFC) the sample to 6 K and applying a magnetic field of 120 mT. (b) Magneto-optical image of trapped magnetic flux in a sample field-cooled (FC) to 6 K in an external magnetic field of 120 mT. (c) Magneto-optical image of trapped magnetic flux in a sample FC to 32 K in an external magnetic field of 120 mT. (d) Current stream lines calculated for c that illustrate the uniform current distribution that circulates inside the bulk even near Tc...... 46

transport magnetization Figure 4.10 - Jc (symbols) and Jc (solid lines) as a function of applied magnetic field at 4.2 K for the K-doped wire compared to other round, untextured, Fe- based superconducting wires. Sm1111 wire is from reference45 and FeSe wire is from reference.82 Inset is an SEM image of the K-doped mono-core wire showing the round cross section with Ag and Cu sheaths...... 47

xiii

response was measured at 4.2 K on a 4 cm pieces of wire with voltage taps approximately 1 cm apart. Measurements were made with increasing current...... 47

Figure 5.1 - Remanent magnetization at 0 T applied field as a function of maximum applied field for (a) wires and (c) tapes. Derivative of data in a and c is presented in (b) and (d) respectively...... 51

Figure 5.2 – Normalized moment as a function of temperature for K-doped Ba122 (a) wires and (b) tapes heat treated at various temperatures. The magnetic response was evaluated by warming above Tc after zero field cooling to 5 K and applying a field of 2 mT parallel to the sample’s length. (c) Tc as a function of heat treatment temperature for onset wires and tapes defined at the onset of diamagnetism (Tc ), 0.1 normalized moment 10% 90% (Tc ), and 0.9 normalized moment (Tc ). (d) Global critical current density at 5 K in self-field as a function of heat treatment for wires and tapes. Inset is global critical current density as a function of inverse grain size for the wires...... 53

Figure 5.3 – Scanning electron microscope images of cleaved superconductor wire cores heat treated at various temperatures as indicated in top left corner of images. Inset in top left image is an optical microscopy image of a polished cross section heat treated at 900 ºC...... 54

Figure 6.1- (a) Normalized Tc as a function of doping level to show the sensitivity of Tc 104–106 to doping level. (b) Normalized moment as a function of T/Tc after zero-field- cooling, applying a small magnetic field, and warming above Tc for round PIT wires with similar grain size...... 60

Figure 6.2 - Remanent magnetization (MR) as a function of increasing maximum applied magnetic field (Hmax) at 5 K for (a) Co-doped Ba122 bulks and (b) K-doped Ba122 bulks sintered at different heat treatment temperatures. Insets are the derivatives of the remanent magnetization as a function of the maximum applied field. Solid arrows indicate location of Hp1 for each sample and hollow arrows indicate location of Hp2 for each sample...... 61

Figure 6.3 - Remanent magnetization (MR) as a function of increasing maximum applied magnetic field (Hmax) at 5 K for powders of the sample heat treated at 600 °C ground to different powder sizes as indicated...... 62

global Figure 6.4 – Jc at self-field for K-doped bulk samples as a function of heat treatment calculated from Hp1 from remanent magnetization MR(Hmax) data and ΔM from magnetization hysteresis M(Happ) data...... 63

xiv

Figure 6.5 – Transport critical current density as a function of magnetic field applied perpendicular and parallel to a wire. Arrows indicate increasing and decreasing applied fields. Schematic of grains and field lines (indicated by arrow arrays) with respect to current (I) and the Lorentz force (F) for the two Ic(Hμ0) measurements...... 64

Figure 6.6 – Magnetic moment as a function of applied field for several different maximum applied fields of a piece of bulk K-doped BaFe2As2 heat treated at 600 °C. Inset is an expanded view of the area indicated by the dashed box...... 65

Figure 7.1 - Magnetic response of (Ba0.6K0.4)Fe2As2, (Ba0.4K0.6)Fe2As2, and Ba(Fe0.92Co0.08)2As2 superconductors: (a) volumetric magnetic susceptibility as a function of temperature after Zero Field Cooling (ZFC) to 5 K, applying 2 mT, and warming the sample; (b) Critical current density (Jc) as a function of applied magnetic field calculated from magnetization measurements at 4.2 K...... 70

Figure 7.2 - A 3-D atom-probe tomographic reconstruction of: (a) (Ba0.6K0.4)Fe2As2; (b) (Ba0.4K0.6)Fe2As2; and (c) Ba(Fe0.92Co0.08)2As2 superconductors. Oxygen atoms are in blue and Ba atoms are in orange, other elements are excluded for a clear display of grain boundary segregation. Each dot represents a single atom, but not to scale...... 72

Figure 7.3 - A 3-D atom-probe tomographic reconstruction of: (a) (Ba0.6K0.4)Fe2As2; (b) (Ba0.4K0.6)Fe2As2; and (c), (d) Ba(Fe0.92Co0.08)2As2 superconductors. Oxygen atoms are in blue and Ba atoms are in orange, other elements are excluded for a clear display of grain boundary segregation. Each dot represents a single atom, but not to scale ...... 74

Figure 8.1 - (a) Light microscopy image of a polished surface of the disk-shaped K- doped Ba122 bulk sample (10 mm diameter and 3.7 mm thick). (b), (c) Remanent (Happ = 0) magneto optical images at (b) 11 K and (c) 20 K for the sample field-cooled under 120 mT. The images show macroscopically uniform trapped field gradient at the perimeter. The white contrast in (b) and (c) corresponds to a high flux density perpendicular to the sample surface...... 80

Figure 8.2 - Trapped field as a function of increasing temperature for the bulk sample stack that was field-cooled magnetized at 5 K. Simplified schematic of sample and Hall probe arrangement...... 80

Figure 8.3 - Magnetic hysteresis loop obtained at 5 K. The sample was zero-field cooled to 5 K and the flux density inside the sample stack (at H2) was recorded as a function of increasing and decreasing external field. The inset shows that the hysteresis loop remains open beyond our maximum applied field of 8 T...... 81

Figure 8.4 - Trapped magnetic field magnet creep at H2. (a) Time dependence of trapped field at 5 K. (b) Normalized magnetic field creep as a function of time at 5, 10, and 20 K...... 82

Figure 8.5 - Comparison between K-doped Ba122 and MgB2. (a) Critical current density 42 138 137 vs. applied magnetic field for Ba122, undoped MgB2, and C-doped MgB2 bulks. Dotted lines are extrapolated data. (b) Maximum trapped field vs. radius for K-doped Ba122 and MgB2 polycrystalline bulks calculated from the data in (a) for an infinite thickness cylinder...... 84

Figure 9.1 – Magnetic response of K-doped BaFe2As2 with extra K additions. (a) Volumetric susceptibility as a function of increasing temperature after zero field and then applying 2 mT. (b) Bulk magnetization critical current density as a function of applied field...... 90

Figure 9.2 – Magnetic response of K-doped BaFe2As2 samples with optimal and overdoping of K. (a) Volumetric susceptibility as a function of increasing temperature after zero field and then applying 2 mT. (b) Bulk magnetization critical current density as a function of applied field...... 90

Figure 9.3 – Magnetic response of K-doped BaFe2As2 samples with addition of other dopants. (a) Volumetric susceptibility as a function of increasing temperature after zero field and then applying 2 mT. (b) Bulk magnetization critical current density as a function of applied field...... 91

Figure 9.4 – Magnetic response of K-doped BaFe2As2 samples with impurity additions. (a) Volumetric susceptibility as a function of increasing temperature (b) Bulk magnetization critical current density as a function of applied field...... 91

Figure C.1 – Images of transverse cross section of various wires with different filament counts and sizes...... 100

Figure C.2 – Volumetric susceptibility of K-doped BaFe2As2 wires with various filament counts as a function of increasing temperature after cooling in zero field and then applying 2 mT...... 101

Figure C.3 – (left) Transport critical current density as a function of applied field and (right) corresponding n-values calculated from the I-V curves...... 101

xvi

LIST OF SYMBOLS, ACRONYMS, AND THEIR MEANINGS a Arbritrary length or constant e The charge of an electron,

A Vector potential, Ec Critical electric field Energy dispersive x-ray A Geometrical factor EDS 0 spectroscopy a0 Lattice parameter F Free energy AC Alternating current FBS Iron based superconductor AE Alkaline earth FIB Focused-ion beam (Alkaline earth)Fe As AE122 2 2 F Normal state free energy compound n0 AP Ambient pressure FP Pinning force APT Atom probe tomography FSU Florida state university American society of ASME GB Grain boundary mechanical engineers at. Atomic GE General electric company Standard atmosphere (unit of atm GM General motors company pressure) Arbitrary length or constant or b H Magnetic field local flux density B Magnetic flux density ħ Reduced Planck’s constant

Ba122 BaFe2As2 h The local magnetic field Bardeen, Cooper, and BCS H// Parallel magnetic field Schrieffer ┴ Bi-2212 Bi2Sr2CaCu2O8 H Perpendicular magnetic field

10 Hc2 defined at intercept of ρ Bi-2223 Bi2Sr2Ca2Cu3O10+x H and 10% of ρ above Tc H defined at intercept of ρ Btrapped Trapped magnetic flux density H50 c2 and 50% of ρ above Tc

90 Hc2 defined at intercept of ρ Bx Local magnetic field profile H and 90% of ρ above Tc

C Arbitrary length or constant Happ Applied magnetic field c Arbitrary length or constant Hc1 Lower critical magnetic field

CIP Cold isostatic press Hc2 Upper critical magnetic field Critical grain boundary Co122 Ba(Fe Co ) As H 1-x x 2 2 GB matching magnetic field D Demagnetization factor HIP Hot isostatic press d Diameter Hirr Irreversibility magnetic field Maximum applied magnetic E Electric field H max field

xvii

H at peak in derivative of max Magnetic property Hp1 remanent magnetization due MPMS global measurement system to MR

Hmax at peak in derivative of Hp1 remanent magnetization due mR Remanent magnetic moment local to MR

HV Vickers hardness MR Remanent magnetization Global (intergranular) I Current m global R remanent magnetic moment Global (intergranular) I Critical current M global c R remanent magnetization Local (intragranular) ID Inner diameter m local R remanent magnetic moment Local (intragranular) j Current density M local R remanent magnetization Grain boundary critical JcBGB MSDS Material safety data sheet current density Intergranular critical current Mechanically activated, self- J bulk MSR c density sustaining reaction Grain boundary critical J gb n n-value c current density Intergranular critical current National high magnetic field J global NHMFL c density laboratory Critical current density Charge carrier number J magnetization derived from magnetization n c p density measurments Critical current density transport Jc derived from transport NSF National science foundation measurments K122 (Ba1-xKx)Fe2As2 nv Flux vortex number density

KC Fracture toughness OD Outer diameter Transition energy of an Occupational safety and K electron going from 2p OSHA α1 3/2 health administration orbital to 1s orbital LED Light emitting diode P Loading force Low temperature laser LTLSM P Power scanning microscopy m Magnetic moment or mass PIT Powder in tube MO Magneto optical PLD Pulsed laser deposition Personal protective MOI Magneto optical imaging PPE equipment

xviii r Radius wt. Weight R Resistance x Arbitrary length or constant RE Rare earth XRD X-ray diffraction (Rare earth)FeAsO RE-1111 1-x y Arbitrary length or constent compound REBCO Rare-earth barium cuprate YBCO YBa2Cu3O7-x SDW Spin density wave z Atomic number SEM Scanning electron microscopy ZFC Zero field cooled SF Self field α Arbitrary coefficient

Sm1111 SmFeAsO1-x β Arbitrary coefficient Superconducting quantum SQUID γ Anisotropy interface device

Sr122 SrFe2As2 θ Arbitrary angle 001 tilt grain boundary STO SrTiO θ 3 GB misorientation angle T Temperature κ Ginzburg-Landau parameter, critical temperature Tc (superconducting transition λ Penetration depth temperature)

Tc taken at 10% of the 10 Tc normalized diamagnetic μ0 Permeability of free space moment Tc taken at 90% of the 90 Tc normalized diamagnetic νϕ Vortex velocity moment Maximum superconducting T max ξ Coherence length c transition temperature Tc taken at 0.1% of the onset Tc normalized diamagnetic ρ Resistivity moment Transmission electron TEM Φ Magnetic flux quantum microscopy 0 TOF Time-of-flight χ vol Volumetric susceptibility V Voltage or volume χ Volumetric susceptibility Vibrating sample VSM Ψ Order parameter magnetometer

xix

ABSTRACT

After nearly seven years of research effort since the discovery of iron-based superconductors,

wires and tapes of K-doped BaFe2As2 have finally been developed by the inexpensive and scalable powder-in-tube technique with critical current densities reaching over 0.1 MAcm-2 at 4.2 K. Such progress relies heavily on the development of synthesis techniques that eliminate cracks and secondary phases. High energy ball milling, during which mechanochemical reactions take place, proves to be effective in producing high quality bulk material. The consolidation of high quality powders under high pressure produces bulk material with a fine grain microstructure and

surprisingly high intergranular current density. We explore the dependence of doped BaFe2As2 superconducting properties on sintering temperature in bulks, wires, and tapes to further optimize these materials and find that grain boundaries continue to act as weak-links, effectively blocking current, and limiting the intergranular critical current density in these materials. However, evidence of composition variation and impurity segregation across grain-boundaries suggests that the weak-linked behavior may still be of an extrinsic nature. Despite the current limiting effects of these weak-links, transport current is high enough in our fine grain material to demonstrate the first > 1 T magnet made out of an iron-based superconductor. These results provide a positive outlook for the potential future use of these materials to produce high field magnets.

CHAPTER ONE

INTRODUCTION

The first chapter of this dissertation provides a background and introduction to iron based superconductors, and to superconductivity in general. Chapter 2 provides an in-depth discussion of experimental techniques used in this work. The remaining chapters are presented as stand- alone publications, including their own, less in-depth, introduction and experimental sections. Chapters 3, 4, and 7 are published works and as such have gone through a peer review process. The remaining chapters have not yet been published or reviewed by peers.

Most modern day superconducting magnets are made using the low temperature superconductors

NbTi and Nb3Sn. These superconductors require low temperature operation (typically 4.2 K) and are typically limited to fields below 16 T. Cuprate superconductors can operate at much higher temperatures and magnetic fields but are limited by high anisotropy and grain boundaries that act as weak-links by blocking current transport. As a result, the cuprate superconductors require costly coated conductor technology that limits its use to a few niche applications.

In 2008 superconductivity was discovered in a new family of superconductors known as Fe- based superconductors (FBS)1. Since then, the FBS ‘family’ has grown to include dozens of

compounds. Of those compounds, doped BaFe2As2 has sufficient properties to make it appealing

for applications, including inexpensive constituents, superconducting transition temperatures (Tc) 2 up to 38 K, upper critical fields (Hc2) over 90 T, and very high intrinsic critical current densities over 1 MAcm-2 (4.2 K, 10 T).3,4 In addition, there is a high versatility in how the superconducting properties can arise and be altered with doping and pressure, providing a unique opportunity to probe and elucidate basic physical phenomena, such as superconductivity and granularity in high temperature superconductors, which continues to be some of the biggest mysteries in condensed matter science.

For applications to be realized, the bulk (global) intergranular critical current density as a global function of field (Jc (H)) must be increased. In 2008, we began exploring synthesis routes and studying the phase relations in FBS materials to eliminate secondary phases that block current. By 2009, weak-link behavior in Co-doped BaFe2As2 was confirmed by measuring the

critical current across grain boundaries in bi-crystals.5 In 2011, we measured surprisingly high 6 critical current density in polycrystalline (bulk) K-doped BaFe2As2, providing motivation for further studies to clarify the extent of intrinsic and extrinsic contributions to weak-link behavior in this material.

global As a result, we designed experiments to further optimize Jc (H), and to understand the weak- link behavior that limits it. This thesis represents an accumulation of our results in which we use processing tools and physical characterization to optimize bulk properties and clarify the extent global to which this material may benefit society. We found that Jc (H) varies significantly with dopant, grain size, and grain boundary composition, and we demonstrated that the best material has properties sufficient to make a bulk superconducting magnet. This thesis describes these findings in further detail.

1.1 – Background The physical properties of these FBS materials have been of great interest since they show unconventional magnetic and superconducting properties and contain Fe, which was long believed to be intrinsically detrimental to superconductivity because Fe is magnetic, and magnetism and superconductivity are typically in competition. While properties and structure can vary considerably for FBS materials, commonalities exist. They are layered materials made up of FeAs (or FeSe) planes sandwiched between layers of other atoms with varying complexities. These FeAs layers are where the active superconducting electron pairs are expected to be most prevalent.7,8 For many of these compounds, superconductivity arises when carriers are doped into the system either chemically, by substituting atoms, or physically by applying pressure.

Currently there are several families of pnictide superconductors that have been discovered. The (RE)FeAsO (RE = rare earth) family, also known as RE-1111, where rare earths include La, Pr,

Nd, Sm, or Gd, is well explored due to the high Tc that is obtainable. The substitution of F for O 9 is employed to get the maximum Tc of 55 K in Sm1111 material. The volatility of O and F and the high stability of oxide phases that tend to exist in Re-1111 samples have provided a significant challenge in producing clean bulk material with robust superconducting 4,10–13 properties. (AE)Fe2As2 is another family of superconductors referred to here as AE122

(AE = alkaline earth) with AE = Sr, Ba, and Ca. Superconductivity can be induced in this family by doping any of the three atomic sites which makes it unique among superconductors. Substituting in the Fe site Co or Ni results in electron doping of the active FeAs layers. Electron doping in the non-active AE layer can also be achieved by substituting RE elements on the AE site (ie La for Sr, Ba, or Ca). Hole doping in the non-active AE layer can be achieved by substituting on the AE site an alkali metal to induce superconductivity. Chemical pressure can induce superconductivity by doping As with a smaller isovalent atom (P). The wide diversity of ways to vary the Fermi surface in AE122 makes this a very interesting system in which to study superconductivity. The lack of volatile elements like O and F makes synthesis easier and the absence of lighter elements with low z-numbers (like oxygen) makes characterization much simpler because analytical techniques like electron microscopy and x-ray spectroscopy are better suited for studying heavier elements. There have been many studies to understand the physical differences in superconducting properties between these various modes of doping. However, investigations into the difference between grain boundary (GB) properties are lacking. In particular, the only bicrystal studies have been on Co-doped AE122,5,14 while the highest

global 6,15 intergranular critical current density Jc is reported in K-doped AE122. Sakagami et al. reported robust GB critical current density in 24 degree [001] tilt bicrystals of P-doped Ba122 at 16 BGB self-field and 4.2 K. Figure 1.1 shows a comparison of transport across GBs (Jc ) vs GB 17 misorientation angle (θGB) of the best Ba122 bicrystals compared to YBCO bicrystals. These data suggest that GBs act as intrinsic weak-links although high angle GBs are not as detrimental BGB to Jc as in YBCO. However, despite these findings, unexpectedly high Jc was later observed in K-doped Ba122 wires6 suggesting that polycrystalline Ba122 material may still be useful for global applications, although further increase of Jc by an order of magnitude and better in field global Jc will be required before applications can be seriously considered. Given that these materials are still very much in their research infancy, we are continuing to study them not just for the scientific merit of doing so, but also for the very real possibility that ways to circumvent the weak-link problem may exist beyond the expensive coated conductor technology that has been developed for YBCO and is still not perfected.

BGB Figure 1.1 - Dependence of the critical current density across GBs Jc as a function of 14 16 the [001] tilt misorientation angle for Co-doped BaFe2As2, P-doped BaFe2As2, and YBCO.17

Figure 1.2 - Upper critical field as a function of temperature for various superconductors.18

Figure 1.2 shows a comparison of Hc2(T) between FBS and the current most common 18 superconductors used in industry. Low temperature superconductors and MgB2 have low

Hc2(T) which limits high field applications. Clearly there is an application potential for Fe-based

superconductors at high-field and low temperature. Although Hc2 and Tc may not be as high as in the cuprates, the low anisotropy (~1.2 for Ba122 vs. >7 for YBCO) is advantageous and since

the grains are less weakly linked, engineering the pnictide conductor is considerably less involved for wires and textured thin films.19,20

1.2 – Parallels Between FBS and Cuprates The iron-arsenic family of superconductors share many parallels with cuprate high temperature superconductors. Sm1111 is a layered material that has the ZrCuSiAs type crystal structure belonging to the tetragonal P4/nmm space group. Ba122 has a closely related ThCr2Si2 type crystal structure. The layered structures of Sm1111 and Ba122 both have covalent bonding in the FeAs layers with interlayer ionic bonding. Sm1111 consists of alternating layers of 2- 2+ 2+ (Fe2As2) and (Sm2O2) while Ba122 swaps out the (Sm2O2) layer for a single divalent atom (Ba2+). This yields a highly two dimensional electronic structure with the FeAs electron carrier layer in the pnictide materials resembling the CuO layer in the cuprates. Superconductivity emerges when the parent compound is doped with holes or electrons to suppress the magnetic order, or when pressure from physical or chemical stress results in lattice distortions that suppress the spin density wave (SDW).21 The structural similarities between the cuprates and ferropnictides, as well as the emergence of high temperature superconductivity through suppression of the SDW, suggest that the cuprates and pnictides share some of the same fundamental ingredients required for high temperature superconductivity to emerge.

One of the main obstacles associated with developing high temperature superconductor (HTS) devices is the inability to transmit current through high angle grain boundaries.12,22 As a result, applications have been limited and rely on expensive processing for HTS requiring techniques in which preferential texturing is used to minimize GB misorientation angles. Applications for superconductors still mostly rely on low temperature superconductors for this very reason. A pressing question is why GBs are detrimental for current flow in some superconductors, while in other superconductors they are transparent to current flow, or sometimes even advantageous in

raising Jc as is the case for Nb3Sn. In the BaFe2As2 system there is evidence of both weak link behavior, in which GBs severely limit current flow, as well as robust GB properties, in which significant current can pass across many high angle GBs. The fundamental superconducting properties in the BaFe2As2 system are strongly influenced by the choice of the dopant which gives rise to superconductivity. While there has been much study of single crystal properties as

they vary with chemical doping, bulk properties remain largely unexplored particularly for well- connected material with clean GBs.

1.3 – Current Transport in Bulk Ferropnictides Assessing the intrinsic properties of the bulk ferropnictide material has been a challenge. FeAs impurity phases have been found in most bulk samples synthesized so far. This non- superconducting phase forms before the bulk material and melts at the temperatures required for synthesis. As a result, grain boundaries are typically surrounded by the FeAs wetting phase that blocks current transport along with a Sm2O3 insulating phase typically found in Sm1111 samples. If clean ferropnictides can be produced with grain boundaries that allow for the transport of supercurrent, then round wires with randomly oriented grain boundaries in the material can be produced. If they exhibit intrinsic weak linked behavior between grain boundaries, then texturing may be required to allow material grains to line up with each other to effectively transport current under applied magnetic fields.

Figure 1.3 shows a polycrystalline Ba122 sample and its corresponding magneto optical image (MOI). The MOI technique visualizes how the magnetic flux penetrates into the sample. The flux is excluded from the center of the grains due to the diamagnetism of the superconductor but penetrates into the non-superconducting FeAs wetting phase and/or intrinsically weak-linked grain boundaries. As a result, the difference in magnetic properties is visualized and shows clear electromagnetic granularity. Ba122 epitaxial thin films grown by pulsed laser deposition (PLD) have allowed for current transport to be analyzed across individual grain boundaries that were free of this wetting phase.5 This is realized by growing the material on bicrystal STO substrates with predetermined orientation angles. Lee et al. used low temperature laser scanning microscopy (LTLSM) and MOI to determine that the Ba122 material grain boundaries intrinsically inhibit the transport of current as a function of grain orientation angle. They were able to compare this data to bicrystal experiments done with YBCO 17 to determine that the current transport for Ba122 was not limited as dramatically with increasing misorientation angle. Ba122 allows for higher angle grain boundaries to transmit current than for YBCO before an exponential decrease in Jc.

Kametani et al. used LTLSM coupled with scanning electron microscopy (SEM) to investigate intergrain current flow in randomly oriented polycrystalline Sm1111 samples confirming grain to grain transport was blocked due to cracks and a high density of FeAs wetting phase. 23 Under 0.1 T magnetic field they found the dissipation spots indicated by LTLSM decreased dramatically indicating weak linked behavior, though they were not able to conclude whether it was an indication of intrinsic weak linked behavior of the material’s own grain boundaries or the extrinsic wetting phase. Other groups have recently reported bulk synthesis techniques for Sm1111 with dramatically reduced impurity phases present that still show a dramatic decrease in 11,24 Jc under a weak magnetic field (< 1 T).

There are several reasons why grain boundaries might limit supercurrent. Stoichiometry variations typically manifest themselves in the grain boundary region due to the presence of vacancies and strains in these high energy regions. This may affect the scattering of carriers, thus affecting critical current density.17 Graser et al. modeled the charging of the interface between dislocation cores, the density of which increases as misorientation angle increases, at the grain boundaries.25

1.4 – Characteristic Lengths in Type-II Superconductors Ginzburg and Landau’s phenomenological theory was developed to describe superconductivity in terms of thermodynamics, postulating that superconductivity was associated with a phase transition that must depend on some superconducting order parameter Ψ based on the 26,27 superconducting charge carrier number density np. In dissertations about superconducting materials, it is standard to go somewhat in depth regarding the derivation of Ginzburg-Landau’s equations, derivations from Bardeen, Cooper, and Schrieffer’s (BCS) complex quantum mechanical theory of superconductivity, and later theories concerning the intricacies of vortex physics.28–31 Since the focus of this dissertation is on the synthesis and characterization of select iron-based superconductors, only a brief introduction to theory is given here with emphasis on equations and characteristic lengths derived elsewhere that are pertinent to the description of mesoscopic phenomenon studied herein.

According to Ginzburg-Landau theory, Abriksov worked out that type-II superconductors have a negative surface energy that allows for magnetic flux to penetrate the superconductors in the form of long filaments when the applied magnetic field H ≥ Hcl, where Hcl is the lower critical

field. When Hc1 > H> Hc2, (Hc2 being the upper critical field) the superconductor is in a mixed state in which the superconducting material is divided into normal and superconducting regions. This negative surface energy results because as the magnetic penetration depth (λ) increases, the diamagnetic screening energy is reduced and as the Ginzburg-Landau coherence length (ξ) gets shorter, the recovery of the Ginzburg-Landau order parameter (ψ) occurs more quickly resulting in more condensation energy. This is true for materials with a Ginzburg-Landau parameter, κ = λ/ ξ, greater than 1/√2. To maximize the boundary surface area the superconductor divides into normal and superconducting regions. The normal (non-superconducting) flux containing regions are known as vortices since super current circulates around them. The free energy (F) of a superconductor near Tc according to Ginzburg-landau theory can be expanded in a series of the form:

2 2 2 β 4 1  h 2e  h F = F +αψ + ψ +  ∇ − Aψ + n0 2 4m  i c  8π

Where Fn0 is the normal state free energy, α and β are temperature dependent coefficients, m is the effective mass, e is the charge of an electron, A is the vector potential, ħ is the reduced

Planck’s constant, and h is the local magnetic field.32 Two key characteristic lengths are derived from the above equation. Ψ is zero in the center of a vortex and recovers to its equilibrium value at a characteristic distance ξ from the vortex’s center, given by the following equation:

h 2 ξ = 2mα

This can also be thought of as the characteristic length describing the spatial size of the superconducting charge carries. The characteristic length for the decay of magnetic field in a superconductor is given as:

= 4 || where μ0 is the permeability of free space. This also describes the distance over which screening currents are induced in a superconductor. Vortices typically contain one quantum of magnetic flux. Once vortices penetrate a superconductor, they have the same sign resulting in a repulsive vortex-vortex interaction. Therefore, in a pure, isotropic, superconductor with large κ and at low temperature, vortex lines form a lattice that is typically hexagonal with some field-dependent

lattice parameter a0.

The critical state model33,34 describes the balance between flux pinning and the external magnetic pressure. Changes in an externally applied field result in induced currents in the surface of the superconducting sample. If the induced currents exceed a certain value, flux flows into the sample until the entire sample is penetrated by vortices, and the critical state is achieved. Once in the critical state, external changes in field result in a change in the internal flux gradient. The force balance between induced current and field generated is often used to deduce the critical current density of a superconducting sample from magnetization measurements.

An external current density will incite a nonzero Lorentz force and if the vortices move they will 35 induce an electric field (E = B x νϕ) where νϕ is the vortex velocity. Power (P) is then dissipated since P is proportional to E∙J. To transport large amounts of current without dissipating energy, vortices must be “pinned” down. That is to say, the Lorentz force needs to be

opposed by some pinning force Fp. This is effectively achieved by providing non- superconducting defects with a diameter d ≈ 2 ξ spaced evenly through the material.35–37 In principle, any spatial inhomogeneity that changes the superfluid density locally will act as a pin. Some of the condensation energy lost by the presence of a vortex is regained when the vortex is pinned through a defect rather than through the superconductor where it is free to move. The average pinning force density in a material can be written as:

= ×

The flux vortex number density varies with field and is given as nv=B/Φ0 with each vortex line 2 taking up a finite area given by 1/nv=(a0 sqrt(3))/2, where a0 is the lattice spacing of the flux line lattice. Written in terms of flux density:

2Φ = √3

As field is increased, the number of vortices increases and a0 decreases until the normal vortex cores eventually overlap around Hc2 since a0 approaches ξ and there is no longer a continuous path for supercurrent.

1.5 – Bean’s Critical State Model Bean’s critical state model33 (here referred to as the Bean model) was developed to explain the irreversible magnetization of type 2 superconductors. The Bean model provides a macroscopic theory that fits well with the ideas theorized by Abrikosov.28 The model assumes that there are two superconducting states, perfect diamagnetism or the mixed state in which partial flux penetration has occurred corresponding to a macroscopic critical current density of 0, Jc, or -Jc.

If a sample is exposed to an electromotive force, Jc will be induced in the sample surface. Above the lower critical field (Hc1), flux penetrates into the sample some distance (x), while the center of the sample remains shielded from magnetic flux. By Ampere’s Law, the gradient of flux lines corresponds to the critical current (dB/dx = μ0Jc) (This relies on an assumption that Jc is independent of the applied field). This penetration of vortices continues in increasing field until the entire sample is in the mixed state. This field-dependent penetration depth explains the size- dependent magnetization curves studied in the following chapters.

CHAPTER TWO

EXPERIMENTAL TECHNIQUES

2.1 – Introduction Many experimental tools and techniques were used in this work. A lot of time, effort, and funding were put into safety considerations, as safety was and is a top priority particularly considering the hazards of working with pnictogens and pressure vessels. Reaction pathways were tested and developed as we studied phase relations and worked to eliminate impurity phases that were found to typically wet the grain boundaries of bulk samples, blocking supercurrent and resulting in poor bulk properties. Mechanochemical alloying and low temperature, high pressure, sintering led to bulk samples with robust superconducting properties. Wire and large bulk fabrication was then developed to characterize for application potential.

Samples were passed through a gantlet of tests to characterize them as new reaction paths were developed and tested. Purity was assessed by X-ray crystallography, magnetic property measurements, and scanning electron microscopy with electron diffraction spectroscopy. Superconducting properties were then gaged by magnetic and physical property measurements and compared to single crystals and bulk data reported in the literature. Microstructural characterization gave clues about grain size, impurity phases, and chemical stoichiometry that were useful to explain anomalies or trends in the superconducting properties that we measured. Often times the most powerful tool in the scientists toolbox is collaboration and it is appropriate to recognize the huge amount of effort and expertise provided by other scientists, laboratory assistants, laboratory managers, and technicians.

2.2 – Safety Considerations

A variety of different synthesis pathways are used to make doped BaFe2As2, and other FBS compounds. The starting materials are combined in a glove box and reacted at temperatures up to 1250 °C in sealed ampules. Potential hazards are the toxicity of As, and the possibility of an explosion when reacting the chemicals in a sealed crucible since As sublimes around 617 °C, and will build up 36 atm of gas pressure before condensing into a liquid. Safety plans were developed (see appendix A) before working and a large amount of effort was put into making a safe environment to work with these materials. An ambient pressure tube furnace was used to

heat treat samples in a stainless steel tubes up to 800 °C. In the event of a crucible explosion, the material is contained within the tube and any hazardous gasses are condensed into sand filters. For most of our samples we used a hot isostatic press (HIP). The press allows for up to 200 MPa of Ar pressure to be applied to sealed samples. From a safety standpoint, the HIP is the safest way to synthesize FBS compounds, because the overpressure prevents As from subliming, so the material is always in the solid or liquid state, and the ASME certified pressure vessel is completely contained and vented outside. An Integrated Safety Management plan was also developed and actively implemented and maintained by each individual working in the laboratory (see Appendix B: Integrated Safety Management Plan).

2.3 – Synthesis Techniques Before making wires or tapes of ferropnictides, high purity bulk material needs to be obtainable. The difficulty of synthesizing phase pure materials for Sm1111 compounds comes from the volatility of As, F, and O in the system, as well as the high stability of Sm-oxide phases. High temperatures are required between 900 ºC and 1400 ºC to create the material and extra F and As additions are often used to compensate for material losses during heat treatments.24 Ba122 material is easier to synthesize since there are only 4 elements most of which are easy to contain during a heat treatment due to their low vapor pressure. There are four synthesis techniques for bulk material covered here; ambient pressure multistep synthesis, mechanochemical synthesis, high pressure synthesis, and powder in tube (PIT) processing.

2.3.1 – Ambient pressure solid-state synthesis Ambient pressure synthesis is done by combining stoichiometric combinations of chemicals and mixing or milling them under Ar. The material is then pressed into a pellet and sealed under Ar or vacuum in a quartz or metal ampoule and heat treated. Typically, several different heat treatments are required with intermittent grinding and repressing between to homogenize the material and reduce impurity phases.23,38–40 Low heating rates are used to avoid the sublimation of elementary As into a gas phase that can cause the ampoule to explode. Alternatively, a binary

compound such as FexAsy or BaxAsy is used to avoid working with elementary As, but such material is not available commercially and has to be synthesized as well. The bulk material obtained is typically not very dense since voids exist between the compacted powder particles and the material is not melted. Even with much care, the 36 atm of pressure generated when

arsenic sublimes is a safety hazard and high pressure synthesis is favored as a safety consideration.

2.3.2 – Mechanochemical synthesis Occasionally, when hand grinding the elements to make Ba122 or M-As binary compounds in a mortar and pestle, exothermic, self-sustaining reactions occurred that were activated by heat from friction between the mortar and pestle. These unintentional reactions further increase the hazards and difficulties of synthesizing these materials because the reactions have large heats of formation that heat the powders, which vaporizes As and other constituents. These reactions were exploited and a new synthesis route for FBS was developed that is covered extensively in chapter 3.

2.3.3 – Hot isostatic pressing High pressure synthesis was carried out by hot isostatic pressing (HIP) during the heat treatment. Powder for the HIP is loaded into a steel ampoule with either a Nb or Ta layer between the ampoule and the sample as a diffusion barrier to prevent the sample from reacting with the steel. The steel ampoule is then evacuated of air, welded shut, and put under pressures up to 240 MPa in a cold isostatic press, to remove void space, and then heat treated under pressures up to 200 MPa. This technique has been employed to make PLD targets used for thin film deposition,5,41 high quality bulks,42 and wires6 with high critical current density. The high pressure technique provides a safer route to processing since the external pressure keeps As from vaporizing while heating. This also allows faster ramp rates to maximum temperature without the need for intermittent grinding steps. The down side to high pressure processing is that an unwanted reaction layer between the sample and the Nb or Ta has to be removed after processing. Subsequent regrinding and firing steps are difficult since the sample has to be cut out of its ampoule followed by removing the reaction layer without exposing it to oxygen that may contaminate the volatile non reacted impurity phases present. However, we found that low temperature heat treatments can minimize the reaction layer and allow for several intermittent grinding steps without significant contamination of material.

2.3.4 – Wire fabrication A powder-in-tube (PIT) method of synthesis has been employed to produce ferropnictide wires and tapes.6,43–50 The PIT method remains one of the most economical and straight-forward approaches to make practical superconducting wire in long lengths. Because most HTS are brittle ceramics or intermetallics, wires are made by filling pre-reacted superconducting powder into a sheath of material (typically Ag, Cu, or another conducting alloy). This is then drawn, swaged, or rolled thinner using a succession of dies that each reduce the cross section by 10-20% until a wire of desired thickness is obtained. Monofilament wires can be restacked and put into a sheath to be drawn down further to create multi-filament wires with exotic geometries and very fine filaments (on the order of micrometers) if restacking and drawing are repeated (see appendix C). The incorporation of non-superconducting pinning centers or structural reinforcement when restacking is also possible. After drawing, rolling, and/or swaging is completed, the wires are then heat treated to allow the powders to sinter and recrystallize into a well-connected network of grains. The sheath material not only provides structural support but can also provide a matrix of conductor to safely disperse current in case the superconductor fails. Before worrying too much about strength and stabilization, Jc(H) must be raised to values needed for applications, typically said to be above 0.1 MAcm-2. PIT processed ferropnictide wires have been studied by many authors and are reviewed elsewhere.46,51 Sheath material explored in these works has included Nb, Ag, Ta, and Fe.52,53 X-ray diffraction analysis of samples indicate impurity phases were present in these materials for all samples and weak links such as secondary phases, voids, or cracks were observed by scanning electron microscopy. Of the four sheath materials mentioned, Ag has been demonstrated to be the best since it does not react chemically with the FBS materials.44,53 For our wires, Ag was chosen because of its ductile nature and chemical inertness to FBS. In addition, we use a Cu or stainless steel tube to provide additional stabilization, strength, and to prevent oxygen from diffusing through the Ag and into the FBS during the heat treatments.

2.3.5 – Large bulk fabrication Large bulk samples are needed as targets for pulsed laser deposition (PLD) and for certain applications, such as use as a large bulk magnet (see Chapter 8). For PLD targets, dense homogenous material is needed, but superconducting connectivity between individual grains is

of little concern since the material will be ablated and re-deposited onto a single crystal substrate. Therefore, high-temperature melt processing is preferred to obtain density near 99% with large grains. For use as a bulk magnet, larger versions of the PIT samples mentioned above were manufactured and heat treated.

For PLD targets, typically 14 g samples were made. Because the SPEX milling vessels we use are limited to 10 g of material, multiple batches of powder need to be milled and combined if all the elements are combined and processed by mechanochemical synthesis. Instead, about 7 g of

Ba3As2 precursors were made by mechanochemical synthesis (see chapter 3), and then they were mixed with the rest of the elements in a mortar and pestle. This was done to obtain a fine- powder Ba-source allowing for better, more even, distribution of elements than is obtainable when working with Ba metal alone, which is soft and not readily made into a fine powder. The powders were wrapped with Nb and sealed in stainless steel (as described in section 2.3.3) and the material was heat treated for 12 hours at 1120°C then ramped down at 4 °C/h and held for 20 hours at 900 °C in a hot isostatic press (HIP) (AIP, Columbus, OH) at 193 MPa to melt the chemicals into a liquid solution. This procedure results in a non-ideal rectangular target geometry, but the material can be mechanically ground into an acceptable shape. HIPping was used to produce dense targets and as a safety precaution to keep the As from subliming. Earlier targets using Ba metal as the Ba-source showed a separation between Ba-metal and FeAs liquid phases during heat treatment due to gravity, and would therefore be Ba-rich on one side of the target and Fe-As rich on the other side. Upon exposure to air, the Ba would oxidize, either wrecking the target or bringing significant uncontrollable amounts of oxygen into grown thin films. We discovered that some oxygen addition was beneficial for Jc in epitaxially PLD grown thin-films due to the incorporation of well disbursed oxide pinning centers.37 Stoichiometric targets were shown to produce films depleted of As that had poor superconducting properties. Attempts to add extra As in the targets resulted in Ba122 + As phase equilibria. Arsenic segregated to the grain boundaries where it oxidized upon exposure to air resulting in cracking of the targets and safety concerns since As-oxides are extremely poisonous. However, the addition of 5% extra As, and 5% less Ba resulted in Ba122 + FexAsy phase equilibria. FeAs phases were observed to be more stable in air resulting in more mechanically robust targets and provided the extra As needed for PLD deposition of high quality films.

Currently available permanent magnets are limited by their magnetization saturation point, and therefore not capable of producing fields much more than 1 Tesla. However, induced currents can be trapped inside of a superconductor to produce magnetic fields (Btrapped) that scale with the size of the bulk. The field trapping ability is then essentially only limited by Jc(H) (Hc2 > 50 T for Ba122) and mechanical strength since the electromagnetic Lorentz force results in stresses proportional to the square of the magnetic field. This is discussed further in chapter 8.

2.4 – Electromagnetic Characterization 2.4.1 – SQUID magnetometry A SQUID magnetometer (Quantum Design: MPMS-XL5s) was used to characterize several superconducting properties. This magnetometer allows magnetization measurements with applied fields up to 5 T. The non-contact magnetization measurements are made by measuring the magnetic induction of two coils as a motor drives the sample between them at about 10 Hz. Bulk samples were typically ground into rectangular slabs that were approximately 0.5 x 1.0 x 4.0 mm. This standard size allowed for quick comparison between samples, easy calculation of volume magnetization, and compatibility with VSM, magneto optical, and physical property measurements. Because of the difficulty of preparing precisely rectangular samples and measurement errors associated with the use of calipers, magnetization calculations of bulk superconductors that are dependent on sample volume have the largest amount of error that can be as much as ± 8 % in some cases for bulks. Wires have much lower volumetric error (≤ 3 %), since their cross sectional area remains relatively constant and was measured by microscopy with better precision while calipers were only used to measure the wire length. Samples were measured with field parallel to the long direction to minimize demagnetization effects. The superconducting magnet was degaussed between measurements, but still had a remanent trapped field of up to ± 7 Oe. The sample itself was used to measure the remanent magnetic field at 5 K and offset it with an applied field for a net field of 0 Oe prior to the measurements. This correction relies on the assumption that the sample is in the Meissner state upon cooling within the magnet’s remanent field, and therefore behaves like a perfect diamagnet.

2.4.2 – M vs T measurements For volumetric susceptibility (χ) vs. T measurements, the samples were cooled in zero applied field to 5 K. At 5 K, the 0 T magnetic moment (m) was measured and offset as mentioned

above. Typically, 20 Oe was then applied and the magnetic moment was measured as a function of increasing temperature at a rate of 0.3 K/min. As temperature is increased, the material

transitions from diamagnetic below Tc to antiferromagnetic above Tc. The broadness of the transition is a pretty good qualitative indicator of sample homogeneity, since a broad transition indicates some portions of the material are more weakly superconducting than others. The magnitude of the measured moment is proportional to the magnetic shielding fraction. For this reason, instead of plotting m vs. T, it is convenient to plot the volumetric susceptibility χ vs T:

4( ) χ = Where m is the magnetic moment in emu, V is the sample volume in cm3 and H is the applied field in Oe. This helps us compare various samples since χ is an intensive property with χ = -1 for a perfect diamagnet. Demagnetization effects were taken into account according to the formula:

χ χ = 1 − ( ∗ χ ) Where χapparent is the measured quantity and D is the demagnetization factor calculated for slab geometry as:

1 = 2 + 1 Where c is the dimension parallel to the field and a is the minor dimension perpendicular to the field. Because of the very small applied field, the antiferromagnetic signal of the parent material

above Tc is negligibly small and any noticeable moment is more likely attributed to magnetic impurity phases. This provides another qualitative assessment of sample quality since additional magnetic impurity phases are not likely benefiting bulk superconductivity.

2.4.3 – Trapped remanent field measurements

For remanent magnetization measurements, a maximum magnetic field (Hmax) is applied and

removed, the remanent magnetic moment (mR) is measured, and this is repeated in progressively

increasing Hmax. Initially, the magnetic field is shielded by the superconductor via induced surface currents and the Meissner effect, but after a high enough field is applied, flux begins to penetrate into the sample where it becomes trapped (flux-pinning effect). For a superconductor with poorly connected grains, flux penetrates preferentially at the grain boundaries. The resulting trapped magnetic flux results in global current flow that produces a global remanent global global magnetic moment (mR ). mR saturates when Hmax is high enough for the magnetic field to completely penetrate into the center of the sample, putting it in the critical state. As the magnetic field is increased further, magnetic flux penetrates into the grains themselves and becomes trapped, which results in current loops on the scale of the grain size that sum to produce a local local remanent magnetic moment (mR ). To quantify the Hmax required for flux to fully penetrate the sample and grains, it is convenient to plot the derivative of mR (dmR) with respect to the derivative log Hmax (dlog Hmax). The peaks in the derivative remanent magnetization can be correlated to different scales of current flow.10,12,23,54 For a weakly linked superconductor, the first peak (at lower field) indicates the applied field at which flux penetrates into the bulk superconductor (Hp1), and the second peak (at higher Hmax) indicates the applied field at which flux penetrates into the grains of the superconductor (Hp2). The location of the peaks can be used global local to calculate Jc and Jc since Hp ~ Jc x (current loop size). Neglecting Jc anisotropy and the demagnetization factor, the Bean model shows for a thin plate with H//c:

21 = where a is the plate thickness. Similarly, assuming spherical grains for the local current:

2.812 = where r is the grain radius. This calculation gives reasonable results for large grained samples but fails for samples with r < λ since the length-scale of shielding currents result in phenomena local that will be discussed in Chapter 6. Unlike mR , which saturates once flux fully penetrates the global grains, mR first increases and then decreases with increasing Hmax. This observed drop in mR as field is increased is beyond the lower critical field of the grains (Hc1G) and can be attributed to the hysteretic behavior of the intergranular critical current density caused by trapped magnetic 54–56 54 flux in the grains. To illustrate this point, figure 2.1 of data from Müller et al. shows mR of

a Bi-2223 tape before and after bending. Before bending, a shoulder is present since flux penetrates on two length scales (mR = mRglobal + mRlocal). After bending, cracks destroy global current and a single transition is observed (mR = mRlocal). If data after bending is subtracted from the initial data, mRglobal can be obtained and a hysteresis of the critical current density

(schematically shown in the inset) is inferred by the presence of a peak mRglobal that then 54–56 decreases as Hmax is increased. This hysteresis is described by several authors and discussed further in Chapter 6.

2.4.4 – VSM magnetometry An Oxford vibrating sample magnetometer (VSM) was used to measure m as a function of T and H up to 14 T. The VSM operates by continuously vibrating the sample between a pair of pick-up coils at a frequency of 55 Hz and amplitude of 1.5 mm. An AC signal is induced in coils and measured by a lock-in device. The VSM allows data to be collected continuously with very little

measurement lag or point to point variation. This allows quick measurements in continuously swept fields. The sensitivity of our device is ~10-6 emu, which is less than the SQUID magnetometer, presenting a trade-off between speed and sensitivity. The higher swept field rates (0.6 T/min for our measurements) than obtainable by the SQUID tends to result in higher measured moments since flux creep (the time dependent exiting of magnetic flux due to thermally driven de-pinning) is less pronounced though too high of ramp rates can result in flux avalanches, though none were observed in our samples.

2.4.5 – M vs. H measurements

Measurements of magnetization as a function of applied fields (Happ) using the tools above are useful to study and understand magnetic materials. Superconductors are nearly perfect diamagnets at Happ lower than Hc1 because Happ induces surface currents that create their own field that opposes the applied one. As a result, Happ < Hc1 exhibits reversible magnetization. At higher Happ, flux penetrates into the material resulting in an irreversible magnetization. This

irreversible magnetization persists up to the irreversibility field (Hirr), at which point bulk

superconductivity has been completely destroyed by the magnetic field. Magnetization Jc can be calculated from the irreversible hysteresis (the difference between the magnetization in increasing and decreasing field) via the extended Bean model for a slab geometry:57

20∆ = (1 − ) 3 where Δm is the difference between the magnetic moment in increasing and decreasing field in emu, V is the volume of the sample, a is the minor sample dimension perpendicular to the field and b is the major sample dimension (perpendicular to a and the field). For cylindrical wires, the following calculation was used:

15∆ = where r is the radius with field applied parallel to the axis of the wire.

The above calculations rely on a few assumptions that are not always valid:

• The self-field is of the sample is much less than the applied field (this assumption is valid for small sample sizes, such as those used in the SQUiD or VSM)

• The lower critical field is zero (this is reasonable for applied fields much higher than Hc1)

• The sample only has three critical current densities (Jc, -Jc, and 0) • Current can travel homogeneously throughout the sample.

For weak-linked materials, Jc is not homogeneously distributed on a single length scale, so the magnetization is proportional to the sum of each current density multiplied by its respective length scales.

2.4.6 – Physical property measurements Physical property measurement systems (PPMS) were used to study select samples in able to see how their physical properties compared to single crystal and bulk material reported in the

literature. Several different magnet systems were used. Resistance (R) vs. (T, Happ) measurements were made in magnet systems with a variable temperature insert, allowing temperature control. Critical current (Ic) vs. (Happ) measurements were made in boiling liquid Helium.

2.4.7 – R vs. (T, H) measurements Resistance (R) was measured using the four-contact method, in which 4 silver wire contacts were affixed to a slab of material using silver paint. Two voltage taps were affixed approximately 1.5- 2 mm from one another near the center of a slab of well-defined geometry (~0.5 x 1 x 4 mm). Current leads were attached to either end using silver paint. The samples were isolated from a sample stage using sapphire and GE varnish. An excitation current of 100-500 μA was applied

and R(T, Happ) was measured. Measurements of R(T) in increasing and decreasing temperature were first done to insure there were not issues with sample heating or thermal lag. 9 T and 16 T Quantum Design PPMS systems were used initially and higher field measurements were made using one of the 35 T resistive magnets at the National High Magnetic Field Laboratory. The

data collected was offset by the average resistance measured well below Tc. This resistance was typically less than 20 nano-ohm. Resistivity was calculated for the slabs according to:

= Where A is the cross sectional area and L is the distance between voltage taps. Uncertainties in this calculation are typically within a few % due mostly to measurements of A and L.

Upper critical fields as a function of temperature Hc2(T) were plotted from resistivity data following three different criteria. First, the linear portion of ρ(T) above Tc was linearly fit for a given H. Then, Hc2(T) was defined for each H at the intercept of ρ(T) and 10 %, 50 %, or 90 % of this line. The three definitions of Hc2(T) are referred to as H10, H50, and H90, respectively. H90 corresponds to where superconducting current paths are first developed within the samples, though resistance is still measured in parallel through most of the sample. H10 corresponds to the

(T, Happ) phase-space under which most of the bulk material is superconducting.

2.4.8 – Ic vs. H measurements Measuring voltage as a function of current is the most direct way to assess the bulk current carrying limitations of a superconductor. The maximum amount of current a superconductor can

carry defines its critical current (Ic). Similarly to resistance measurements, physical contacts have to be made for these measurements. However, because of the much larger currents that have to be applied, very low resistance contacts and stabilizer are needed to prevent heating the area between the voltage taps. To do this, we made PIT wires with metal sheath material that could be soldered to large copper current leads. The voltage taps were typically spaced ~1 cm from one another in the center of 4 cm long samples. The wires were measured with fields perpendicular to the wire current in the maximum Lorentz force direction, and parallel to some wires in the minimum Lorentz force direction. The data was offset by the background voltage -1 (below Ic). Ic was defined using a criterion for the critical electric field (Ec) of 1 μVcm . The

sharpness of the Ic transition can be defined by the n-value (n) that comes from fitting the data with the following function:

() = + + where a and b are free parameters to account for the background. To insure V(I) measurements are occurring at the right temperature and that the transition is not a result of heating from the

current leads, measurements were made in both increasing and decreasing currents to make sure

there was not a hysteresis of Ic that would indicate a change in sample temperature as it was transport measured. Jc is defined as Ic divided by the cross sectional area of the superconducting wire core.

2.4.9 – Magneto optical imaging (MOI)

Magneto optical (MO) imaging is used to image the local field profile Bx produced by magnetization currents induced by applied magnetic fields. MO imaging uses a thin Faraday- active crystal (5 μm thick Bi-doped garnet indicator film for these studies) placed onto the sample surface to image the normal field component produced by magnetization currents induced by magnetic fields applied perpendicular to the sample’s surface. This is achieved because the rotation of light polarization is sensitive to small magnetic fields. The magneto- optical Faraday effect is powerful for characterization since it allows the space-resolved measurements of magnetic flux density distribution in a superconductor on a mesoscopic length scale (> 1 μm). An MO microscope is a regular light microscope with filters and the addition of a cryostat to manipulate sample temperature and a small electromagnet to apply a magnetic field. In addition to allowing direct observation of magnetic flux density distribution in a superconductor in real time as temperature and field are manipulated, the current density can be calculated via Ampere’s law and the electric field can be calculated via Maxwell’s equation. Algorithms for numerical inversion of the Biot-Savart law can give an integral relation between the magnetic self-field and the current density distribution.58 Grain boundaries, inhomogeneity,

cracks and Jc anisotropy give rise to complex current patterns, making MOI well suited for assessing current distribution on a scale down to a few microns. The MOI in this work was carried out by Anatolii Polyanskii.

2.5 – Microstructural Characterization 2.5.1 – Optical imaging Olympus optical microscopes with attached cameras were used for sample preparation, documentation, and many of the measurements in this dissertation. Polarizing filters were used on the incident and reflected light optics when analyzing well-polished samples in able to get good granular contrast to quantify grain size. This is possible because of the anisotropic optical properties of FBSs. The amount of impurity phases, such as FeAs, could also be differentiated

using non-polarized light and then quantified using Adobe Photoshop. Stream Motion software allowed for quick length and area measurements on samples prior to electromagnetic characterization. The error in such measurements is estimated to be around ± 0.001 mm. This is over an order of magnitude less than length measurements using a caliper (typically ± 0.025 mm). All software calibration was checked with a Meiji stage micrometer with 0.01 mm divisions.

2.5.2 – Scanning electron microscopy (SEM) A Zeiss 1540 EsB/XB scanning electron microscope allowed for microstructural investigations beyond the resolution limits of the optical microscope (~1 μm). Samples were first dry-polished using progressively finer silicon carbide paper up to 1200 grit/inch paper. Samples were then

wet polished using a methanol/ 0.05 μm Al2O3 solution. After cleaning the surface with methanol, samples were immediately (within 1 minute) inserted into the SEM to avoid sample contamination. Fresh methanol gave better results since it contained less absorbed moisture. Alternatively, freshly cleaved samples or powder deposited on double-sided carbon tape was imaged. SEM was used to identify impurity phases, cracks, and grain size to compare with electromagnetic characterization. The regular optics were used on cleaved samples to see the topology clearly and a backscattering detector was used on the polished samples to get a clear contrast between phases with different densities. 20 kV was the operating voltage for most phase analysis, while lower values around 10 kV were used to probe surface impurities and enhance the channeling contrast to resolve grain sizes that were less than 1 μm. Typical working distance was between 4 and 5 mm to allow enough room for the back scattering detector to be inserted.

2.5.3 – X-ray diffractometry (XRD) X-ray diffraction was used in this work primarily for phase identification following synthesis and processing steps. X-rays were generated by irradiating Cu metal with high-energy electrons. Typical operating parameters for the source were 40 kV and 30 mA. The XRD patterns were

imaged on a HUBER imaging plate Guinier camera (model 670) using Cu Kα1 radiation with a Ge monochrometer in steps of 0.005° in 2θ. Samples were measured in transmission under vacuum and Mylar or generic transparent tape (Office Max) was used to hold the powders during XRD to minimize sample contamination. For powder x-ray diffraction, if phases are below ~3% volume, their presence is typically hidden in the noise of the measurements making XRD only

suitable for rough estimates of phase assemblage. In addition, amorphous or highly disordered material are not readily identified by XRD measurements since they don’t scatter x-rays constructively.

2.5.4 – Energy-dispersive x-ray spectroscopy (EDS) EDS is an analytical technique that was carried out within the SEM microscopes. An x-ray emission spectrum can be gathered when the electron beam is focused on select parts of the sample. This allows for quantitative chemical characterization of phases down to the size of the spot of excitation, typically ~2-5 μm at 20 kV. The resolution of the EDS depends on the energy of the characteristic x-rays escaping the sample and the density of the material it must pass through. For this reason, quantification of lighter elements like O and C is typically poor, while heavier elements can be quantified within an estimated < 8 atomic % error for our system, which was not calibrated with a FBS standard.

2.6 – Nanostructural Characterization 2.6.1 – Transmission electron microscopy (TEM) TEM is a complex microstructural analysis technique that is widely used by the material science community. In this work, it was used to observe the nanostructure on a finer length scale than obtainable by SEM. Specifically, impurity segregation to grain boundaries on the order of a few nm that might be responsible for blocking current were sought out by TEM.5,59,60 TEM was also used to quantify grain size in some samples and look for texture. A JEOL JEM2011 transmission electron microscope (TEM) operated by Fumitake Kametani was used. The main limitation of TEM in this dissertation is that quantitative chemical analysis by electron energy loss spectroscopy was hindered by the very air-sensitive nature of the thin foil samples that are required for this technique.

2.6.2 – Atom probe tomography (APT) APT is an emerging analytical technique that enables the quantification of atomic chemistry within a material in 3d-space with sub-nm resolution. The basic operating principle is presented schematically in figure 2.2. A focused ion beam is used within an electron microscope to machine out a needle shaped specimen. This specimen is exposed to an electric field and then ions are field evaporated into a position-sensitive mass spectrometer. The time-of-flight and

25 position information gathered by the detector allows for the ion position before evaporation to be back calculated and the mass spectrometer determines the charge to mass ratio of the ion. This technique was carried out at Northwestern University by Yoon-Jun Kim and details are presented in chapter 7.

Figure 2.2 - Schematic of atomic probe microscopy setup. Arrows indicate the direction of evaporated ions.

CHAPTER THREE

MECHANOCHEMICAL SYNTHESIS OF PNICTIDE COMPOUNDS

The following chapter was published in the Institute of Physics’ journal Superconductor Science and Technology.42 Vibrating sample magnetometry measurements were carried out by Jianyi Jiang. Magneto-optical imaging was carried out by Anatolii Polyanskii. The research was directed by Eric Hellstrom and all authors discussed the results and commented on the manuscript.

BaFe2As2 (Ba122) and (Ba0.6K0.4)Fe2As2 (K-doped Ba122) powders were successfully synthesized from the elements using a reaction method that incorporates a mechanochemical reaction using high-impact ball milling42. Mechanically-activated, self-sustaining reactions (MSR) were observed while milling the elements together to form these compounds. After the MSR, the Ba122 phase had formed, the powder had an average grain size < 1 μm, and the material was effectively mixed. X-ray diffraction confirmed Ba122 was the primary phase present after milling. Heat treatment of the K-doped MSR powder at high temperature (1120 °C) and pressure yielded dense samples with high phase purity, but only granular current flow could be visualized by magneto optical imaging. In contrast, a short, low temperature (600 °C), heat treatment at ambient pressure resulted in global current flow throughout the bulk sample even though the density was lower and impurity phases were more prevalent. An optimized heat treatment involving a two-step, low temperature (600 °C), heat treatment of the MSR powder produced bulk material with very high critical current density above 0.1 MAcm-2 at 4.2 K and self-field (SF).

3.1 – Introduction Since their discovery, the ferropnictide materials have spawned great interest in the physics and material science communities for investigating the mechanism of superconductivity. They also show promise for applications4,6,61. Many superconducting devices require long wires or tapes of homogeneous superconducting material with well-connected grains that allow supercurrent to flow through the bulk material. Most of the ferropnictide wires and bulks made to date exhibit extrinsic weak link behavior as a result of low density, cracking, and impurity phases that wet the grain boundaries10,11,13,43,44,47,62,63 in addition to possible intrinsic weak link behavior attributed to 27

grain boundaries5,14. Figure 3.1 is a scanning electron microscopy (SEM) image of a Ba122 bulk prepared by a typical grind-and-react route that shows clearly some of the impurity phases that often wet grain boundaries and block current. Safe and efficient reaction methods are needed to produce phase-pure bulk material that can be used to make superconducting devices.

Figure 3.1 - SEM image of BaFe2As2 made by hand grinding Fe, Ba3As2, and As followed by HIP treatment at 1120 °C. The image shows voids and Fe2As that wets grain boundaries.

The risk of an explosion from the high vapor pressure of As at high temperature and its being poisonous have resulted in reaction methods for ferropnictides that typically involve slow heating, intermediate grinding steps, and/or high pressure synthesis10,11,38,40,43,63–67. Multistep synthesis routes that use binary compounds, such as M-As (M=Ba, Sm, K, Co, Fe, La, or Nd), instead of the elements to prevent the vaporization of As are also common39,59,68, but synthesizing the M-As compounds has the same risks. The origin of the present study was our observation that occasionally when hand grinding the elements to make Ba122 or M-As binary compounds in a mortar and pestle, exothermic, self-sustaining reactions occurred that were activated by heat from friction between the mortar and pestle. These unintentional reactions further increase the hazards and difficulties of synthesizing these materials because the reactions have large heats of formation that heat the powders, which vaporizes As and other constituents.

Igniting a self-sustaining reaction with a mechanical impact, such as with the mortar and pestel, is referred to as a mechanically-activated, self-sustaining reaction (MSR). We realized we could use high-energy ball milling to create controlled MSRs. We initiated MSR reactions in mixtures

of Ba and As, and Sm and As that we subsequently used to synthesize Ba122 or SmFeAsO1-xFx (Sm1111). We advanced from these binary systems and investigated whether we could use MSR reactions to form Ba122 directly from the elements. Direct synthesis of Ba122 by MSR is successful and is reported here. The MSR powders have the desired overall composition, i.e., there is no loss of volatile elements, and the particles have submicron grain sizes that aid in homogeneously mixing the constituents. Having a well-mixed powder with very fine grain size is desirable because it promotes the formation of more homogonous, phase-pure material. Using MSR powders can help retain more volatile elements such as As and K in the final product since it allows for shorter reaction times at lower temperatures than are typically required when using standard solid-state synthesis69,70. In this study we found that the MSR occurred just minutes after starting the milling. The MSR-produced powder was characterized before and after the MSR reaction, and after different heat treatments to form bulk samples of K-doped Ba122.

3.2 – Experimental Details All handling of material and milling was carried out in Ar-atmosphere glove boxes. We used As particles (20 mesh, Alfa Aesar), Ba pieces (~20 mesh, Alfa Aesar), Fe powder (200 mesh, Alfa Aesar) and K chunks (Alfa Aesar) in our experiments. Typically 7 g batches of Ba122 powder

with nominal composition BaFe2As2 and (Ba0.6K0.4)Fe2As2 were synthesized by combining stoichiometric amounts of Ba, Fe, As, and K in a zirconia milling vial (SPEX 6.4 cm diameter, 6.8 cm long) with two 12.7 mm diameter zirconia milling balls. Samples were milled using a high-impact milling machine (SPEX 8000M) for 1 hour inside the Ar glove box. To monitor the exothermic MSR reactions, thermocouples were attached to the surface of a smaller steel milling jar (3.8 cm diameter, 6.6 cm long) loaded with three 9.5 mm diameter steel milling balls and a 3 g batch of the elements. Temperature as a function of milling time was measured for reactions between Ba and As, and the elements to form Ba122 and K-doped Ba122.

After 60 min of milling in the zirconia jar, the powder was removed, and then wrapped with Nb foil and placed into stainless steel ampoules. The ampoules were evacuated, welded shut, compressed with a cold isostatic press (CIP) (AIP Columbus, OH) at 275 MPa to press the

29 powder into a pellet, and then heat treated. Three different heat treatments were used and the amount of powder used to make the pellet was adjusted for the specific heat treatment. A sample with 2 g of powder was heat treated for 12 hours at 1120°C then ramped down at 4 °C/h and held for 20 hours at 900 °C in a hot isostatic press (HIP) (AIP, Columbus, OH) at 193 MPa (sample name = 1120 °C HIP). A sample made with 0.5 g of powder was heated for 10 hours at 600 °C at ambient pressure (AP, 1 bar) (sample name = 600 °C AP). The third sample was made by heat treating 5 g of the powder in the HIP under 192 MPa of pressure at 600 °C for 20 hours. This sample was then re-milled and 1.5 g of powder was heat treated again in the HIP (600 °C, 193 MPa) for 10 hours (sample name = 600 °C HIP). The baseline undoped Ba122 sample shown in figure 1 was made using our previous technique of hand grinding stoichiometric amounts of As,

Fe, and Ba3As2 in a zirconia mortar and pestle for 60 min. No MSR was observed for this sample. This baseline Ba122 sample was placed in Nb foil, encapsulated, CIPped, and then HIPped as described above for the sample heat treated at 1120 °C.

X-ray diffraction (XRD) patterns were recorded by a HUBER imaging plate Guinier camera

(model 670) using Cu Kα1 radiation with a Ge monochrometer in steps of 0.005°. Magnetization of the heat-treated samples was measured by a SQUID magnetometer (Quantum Design: MPMS- XL5s) and a 14 T Oxford vibrating sample magnetometer (VSM) with the magnetic field parallel to the sample’s length. Magneto optical (MO) imaging with a 5 μm thick Bi-doped garnet indicator film placed onto the sample surface was used to image the normal field component produced by magnetization currents induced by magnetic fields applied perpendicular to the sample’s surface. SEM images were taken with a Zeiss SEM at 20 kV.

Figure 3.2 - Milling vial temperature as a function of milling time for three different MSR reactions. The baseline was obtained by milling fully-reacted BaFe2As2 powder.

3.3 – Results Figure 3.2 shows the vial temperature as a function of milling time for powders that have BaAs,

BaFe2As2 and Ba0.6K0.4Fe2As2 stoichiometries. The MSR was observed as the sharp increase in vial temperature. The baseline shows a steady temperature increase due to energy input from impact of the milling balls and due to the ambient temperature rise within the glove box from the

SPEX mill components such as the motor. The XRD patterns of the K-doped Ba122 powder just before the MSR and after 60 minutes of milling are displayed in figure 3.3(a) and 3.3(b), respectively. An SEM image of the milled powder is included in figure 3(b) showing a submicron particle size. The XRD pattern (figure 3.3(b)) shows that K-doped Ba122 is the primary crystalline phase indicating that K-doped BaFe2As2 formed during the MSR reaction. There are minor extra peaks in the pattern due to the tape we used to hold and cover the sample. As shown by XRD (left) and SEM (right) in figure 3.3(c), heating at 1120 °C turned the MSR powder into nearly pure K-doped Ba122 bulk with grain size over 60 μm. A small amount of

Fe+Fe2As (<1 vol%) was identified by energy-dispersive X-ray spectroscopy (EDS). Figures 3.3(d) and 3.3(e) show the XRD patterns and SEM micrographs of samples 600 °C AP and 600 °C HIP, respectively. The XRD patterns show sharp diffraction peaks and only trace minority phase peaks from the FeAs phase. SEM revealed that both samples contained more voids and FeAs phase than observed in the sample heat treated at 1120 °C, but the FeAs existed in the 600 °C samples as discrete particles that did not wet the grain boundaries and accounted for about 2-3 vol% of the sample.

0.2 1120°C HIP 0.0 600°C AP -0.2 600°C HIP -0.4

vol -0.6 χ

-0.8 -1.0 -1.2 5 10 15 20 25 30 35 40 45 Temperature (K) Figure 3.4 - Temperature dependence of the volumetric susceptibility under zero field cooling (ZFC) in an external field of 20 Oe for (Ba0.6K0.4)Fe2As2 samples 1120 ° C HIP, 600 ° C AP, and 600 ° C HIP.

The volumetric susceptibility measurements, shown in figure 3.4, were taken after zero-field- cooling (ZFC) the samples to 5 K and then applying a magnetic field of 20 Oe. All three samples show a sharp transition and nearly perfect diamagnetism. Sample 1120 °C HIP has the sharpest

Table 3.1 - Material properties of superconducting bulks.

J (4.2 K, SF) J (4.2 K, 10 T) Fraction of Tc c c Density Sample -2 -2 3 theoretical (K) (kAcm ) (kAcm ) (g/cm ) density 1120 °C 5.76 0.985 38.7 NA NA HIP (± 0.04) (± 0.007)

4.0 0.68 600 °C AP 37.7 11.0 0.9 (± 0.1) (± 0.02)

5.4 0.92 600 °C HIP 37.1 117.0 8.9 (± 0.1) (± 0.02)

transition, indicating good electromagnetic homogeneity throughout the superconducting bulk, but a slight residual magnetization above Tc from the magnetic impurity phase observed by

SEM. Tc decreases slightly from 38.5 K for sample 1120 °C HIP to around 37 K for the samples heat treated at 600 °C (see table 3.1). Sample 600 °C HIP has a sharper transition than sample

600 °C AP, indicating the two-step, high-pressure synthesis improved the electromagnetic homogeneity.

Figure 3.6 - Magnetic field dependence of the critical current density calculated from magnetization measurements at 4.2, 10, 15, 20, 25, and 30 K for samples (a) 600 ° C AP, and (b) 600 ° C HIP.

MO imaging was used to visualize the local field profile produced by magnetization currents induced by an external magnetic field applied perpendicular to each sample’s surface after zero field cooling and by the remanent magnetization when the field was removed. The MO image in figure 3.5(a) of sample 1120 °C HIP shows flux penetration at the grain boundaries and local regions of partial flux penetration within individual grains produced by granular current flow, but no indication of flux produced by current flowing over the whole sample. Figure 3.5(b) shows the corresponding remanent field. Figures 3.5(c) and 3.5(d) are MO images of sample 600 °C AP showing a nearly uniform rooftop pattern of magnetic flux density produced by global current flow throughout the entire bulk sample after an external magnetic field is applied and then removed, respectively. Figure 3.5(e) shows that sample 600 °C HIP has a much more uniform

rooftop pattern of magnetic flux density and less flux penetration than sample 600 °C AP, indicative of better connectivity. Figure 3.5(f) is the corresponding remanent field produced by the bulk when the field was removed. Figure 3.6 shows the magnetization critical current desity

(Jc) as a function of applied magnetic field for both 600 °C samples. Jc was calculated using the Bean model, using the sample’s bulk dimensions, from magnetization hysteresis loops taken at

-2 several temperatures. The high Jc, which is above 0.1 MAcm (<10 K, SF) for sample 600 °C HIP, is consistent with the highest values reported for ferropnictide wires6.

3.4 – Discussion The literature shows that the starting particle size of the powders being milled affects the time to ignition and the kinetics of the combustive reaction front71. Typical MSR reactions reported in the literature begin with fine powders (>100 mesh) and the time to ignition can vary from minutes to hours depending on the reactants and milling conditions69,71,72. The temperature at which ignition occurs for an MSR is thought to decrease with decreasing particle size and with increasing mixing of these smaller particles71. After milling for some time, the ignition temperature is decreased until it is lower than the local temperature generated by the impact between the milling vial and media, and then the MSR occurs. In addition, milling causes a general increase in powder temperature and creates a high density of defects in the powder that can affect the reaction kinetics of the MSR reaction. The sharp increase in the temperature of the outer surface of the milling vial within 1 to 10 min indicated an exothermic MSR reaction had occurred in doped and undoped Ba122 samples as well as in BaAs in a much shorter time than reported in the literature for most other MSRs69,71,72.

Ba and K are particularly difficult to work with when grinding powders together for normal solid-state reactions. They are difficult to grind into high-purity, fine-grain size powder because both of them are soft and easily contaminated with oxygen. Using MSR alleviates these problems because the MSR, which occurs rapidly using high-impact ball milling, produce a brittle, less reactive intermetallic compound that mixes better during the ball milling. The MSR-

35 produced powder also has finer grain size than can be obtained with hand grinding, so the MSR powder sinters at lower temperatures. We have studied MSR reactions between other electropositive metals and pnictide reactants that are not reported here and based on these results we believe that MSR reactions will occur in other systems, including superconducting ferropnictides, that combine highly electropositive elements like Ba, Sr, K, Sm, Nd with electronegative elements such as As, P, and O. Here too, the MSR reaction should lead to higher phase purity.

The heat treatments were chosen to cover a wide range of processing conditions. The high- pressure and high-temperature reaction used for sample 1120 °C HIP resulted in a material with very high phase purity (< 1 vol% impurities observed by SEM) and very high density (~ 98% dense). This heating schedule was previously developed and optimized by our group for making high quality Ba122 samples from the elements without using MSR powder. The high Tc and sharp transition indicate the superconducting part of the material is uniform in composition. However, no global current flow was observed in this sample because of cracks and impurity phases that segregated to the grain boundaries, a common problem for bulks processed at high temperatures10,13,63,67.

The low reaction temperature (600 °C) was chosen to prevent FeAs liquid phases from forming and to stay below the vaporization temperature of As. Staying below the liquidus line in the Fe- As phase diagram insures these impurities won’t wet the grain boundaries and staying below the vaporization temperature of As keeps residual As from subliming, which could cause an explosion hazard and risk of As escaping the reaction crucible. This low reaction temperature also results in a very small grain size that yields fewer cracks and may improve the intergranular current density by improving the flux vortex dynamics as speculated elsewhere6. We see by MO imaging and magnetization measurements that even though sample 600 °C AP is less dense

(~68% theoretical density) and has a broader temperature dependence of χvol than sample 1120 °C HIP, which can be attributed to an inhomogeneous distribution of K, there is significant intergranular current transport. It is interesting to note that that the 68% theoretical density (table 1) is close to the theoretical density (74 %) obtained by packing monosized, spherical particles together. For sample 600 °C HIP, we used the second milling and second heat treatment to improve homogeneity and HIP processing to increase the density. The improved electromagnetic

36 properties compared to sample 600 °C AP can be seen by the sharper χvol transition (Figure 3.4) and the density was increased to ~92% of the theoretical density.

The MO images in figure 5 clearly show that cracks and grain boundary phases can significantly block current for samples 1120 °C HIP and 600 °C AP. Sample 600 °C HIP was free of cracks and grain-wetting phases and had the highest Jc as shown in figure 3.6 (b) with weak field 5 -2 dependence at high fields and temperatures. Its Jc is above 10 Acm at low fields at ≤ 10 K, which is around the values required for applications. If Jc can be increased by a factor of 10 in field, K-doped Ba122 would be useful for high field applications in the temperature range

obtainable by cryo-coolers and liquid H2.

3.5 – Conclusions MSR reactions that can occur when hand grinding elements to form Ba122 can be hazardous. We studied controlled MSR reactions to form Ba122 and K-doped Ba122 using high-impact ball milling. Pre-reacting mixtures of the elements by MSR yields submicron-size, as-milled powder with Ba122 as the primary crystalline phase. After a high temperature HIP heat treatment (1120 °C), the K-doped Ba122 bulk materials made from MSR powder have high density (~98.5% dense) and higher phase purity (< 1% secondary phase by SEM) than can be obtained by HIPping hand-ground powders, but they did not carry global supercurrent due to cracks and the impurity phase wetting, and blocking, the grain boundaries. A low temperature reaction at ambient pressure and 600 °C, which is well below the melting point of the FeAs impurity phase, resulted in bulk material that could carry over 10 kAcm-2 (4.2 K, SF) despite being less dense (~ 68%) and having more impurity phases (>1 vol%) than the material made from the same MSR powder by HIPping at 1120 °C. Using a two-step HIP heat treatment at 600 °C resulted in a material that was ~92% dense, with 1-3 vol% impurity phases and good connectivity, with Jc > 105 Acm-2 at temperatures ≤ 10 K in SF. We believe our new MSR synthesis route can be used to safely synthesize other ferropnictides and will result in bulk materials that have higher phase purity with fewer reaction steps than traditional reaction pathways.

CHAPTER FOUR

HIGH INTERGRANULAR CURRENT DENSITY IN FINE GRAIN FERROPNICTIDES

The following chapter was published in the journal Nature Materials.6 Chiara Tarantini carried out the high-field resistivity measurements and helped prepare the manuscript. Jianyi Jiang carried out vibrating sample magnetometer measurements and helped design the experiments. Fumitake Kametani carried out transmission electron microscopy measurements. ‘Anatolii Polyanskii carried out magneto-optical imaging. David Larbalestier and Eric Hellstrom directed the research and contributed to manuscript preparation. All authors discussed the results and implications and commented on the manuscript.

The K- and Co-doped BaFe2As2 (Ba122) superconducting compounds are potentially useful for applications because they have upper critical fields (Hc2) of well over 50 T, Hc2 anisotropy γ< 2, -2 2,4,20,41 and thin film critical current densities Jc exceeding 1 MAcm at 4.2 K. However, thin- film bicrystals of Co-doped Ba122 clearly exhibit weak link behavior for [001] tilt misorientations of more than about 5°, suggesting that textured substrates would be needed for applications, as in the cuprates5,14. Here we present a contrary and very much more positive

result in which untextured polycrystalline (Ba0.6K0.4)Fe2As2 bulks and round wires with high grain boundary density have transport critical current densities well over 0.1 MAcm-2 (SF, 4.2 K), more than 10 times higher than that of any other round untextured ferropnictide wire and 4-5 times higher than the best textured flat wire73. The enhanced grain connectivity is ascribed to their much improved phase purity and to the enhanced vortex stiffness of this low-anisotropy compound (γ∼ 1-2) compared to YBa2Cu3O7-x (γ∼ 5).

4.1 – Introduction The tendency of grain boundaries of the high temperature superconducting cuprates like

YBa2Cu3O7-x to be weak linked has been the major impediment to producing wires needed for their application17,74. In the search for new classes of superconductors that might displace Nb-Ti and Nb3Sn, the new Fe-based superconductors also demand attention, even though they seem to share similar non-superconducting parent phases to each other and have low carrier densities12. global Bulk ferropnictide materials made thus far exhibit low global Jc (Jc ) values, some of which

can be ascribed to extrinsic factors such as their less than full density, the prevalence of grain- 11,23,43–45,52 global boundary-wetting phases, and cracking . As a result, Jc of randomly oriented ferropnictide wires are typically well below 0.01 MAcm-2 43,75,76, although a recent textured K- 73 -2 doped SrFe2As2 (Sr122) tape reports about 0.025 MAcm at self-field (SF). These values are local one to two orders of magnitude less than the local intragrain critical current density Jc measured in single crystals77–79. A potentially serious intrinsic problem is the weak link behavior of grain boundaries (GBs) in Ba(Fe1-xCox)2As2 similar to that observed in the cuprates

YBa2Cu3O7-x (YBCO) and Bi2Sr2CaCu2O8 (Bi-2212). However the intergrain critical current gb density across [001] tilt misoriented grain boundaries (Jc ) decreases less rapidly with increasing grain misorientation angle than in YBCO5,12,14. The generality of this result is still unclear though because thin film bicrystals of ferropnictides have so far only been grown for one structure (Ba122) in only the Co-doped variant due to the difficulty of maintaining proper composition control during deposition. Therefore, the study of polycrystalline samples is still of great importance. Here we report a surprisingly positive result with (Ba0.6K0.4)Fe2As2 bulks and wires global made by careful low temperature synthesis showing much higher Jc than in Co-doped Ba122 bulks. The very fine grain size (~200 nm), which is comparable to or smaller than the penetration

depth, and the low Hc2 anisotropy provide a basis for high vortex stiffness. The enhanced phase purity at the grain boundaries and low anisotropy appear to enable transport critical current densities that are high enough to be interesting for applications.

4.2 – Experimental Details K-doped Ba122 polycrystals were fabricated by a reaction pathway that results in a more complete reaction, minimizing the formation of current blocking secondary phases, like FeAs, that tend to wet the grain boundaries when higher temperatures are used. Keeping the reaction temperature well below the melting temperature of Fe-As phases insures that the impurities present do not wet grain boundaries where they would severely block the flow of supercurrent. An additional benefit of low temperature reactions is that the grain size of the Ba122 phase is very fine, ~200 nm. Moreover, the bulk forms were easily powdered and made into a wire by the powder-in-tube technique, lengths of which were reacted and characterized.

The elements were first mixed to obtain nominal composition (Ba0.6K0.4)Fe2As2 or

Ba(Fe1.84Co0.8)2As2 and then ball milled for 1 hour. During the milling, an exothermic reaction

occurred, partially reacting the material to the Ba122 phase. The ball milled material was wrapped with Nb foil and placed in a stainless steel ampoule that was evacuated, welded shut, compressed into a pellet with a cold isostatic press (CIP) at 275 MPa, and then heat treated in a hot isostatic press (HIP) under 192 MPa of pressure at 600 °C for 20 hours. The material was remilled and heat treated again as above for 10 hours to obtain a more homogenous bulk. For the wires, the bulk material was milled after the previous two steps and packed into a Ag tube (6.35 mm OD, 4.35 mm ID). The tube ends were plugged, swaged and welded shut. The tube was then groove rolled, followed by drawing to a 0.8 mm OD wire. Pieces of the Ag-clad wire were sealed in Cu tubing (1.57 mm OD, 0.86 mm ID) under vacuum by welding the ends shut. The Cu tubing was then groove rolled to an OD of ~1.35 mm. The Cu/Ag clad wires were then compressed in a CIP under 2 GPa pressure and heat treated for 10 hours at 600 °C in the HIP, as above.

Careful studies of electromagnetic granularity and the broader superconducting properties were

made by SQUID magnetometry for Tc, vibrating sample magnetometry (VSM) to measure the global magnetic moment in high fields, magneto-optical imaging to measure local granularity on scales of a few μm, as well as transport critical current measurements. Microstructures were examined at multiple length scales in a Zeiss 1540 EsB/XB scanning electron microscope (SEM) and a JEOL JEM2011 transmission electron microscope (TEM).

0.0 K-doped Ba-122 wire K-doped Ba-122 bulk

-0.5

χ χ χ χ

-1.0

0 10 20 30 40 50 Temperature (K)

4.3 – Results a 0.08 0.6 RRR~6.8 0.4 ) 0.06 35T 0.2 32.5T 30T cm 0.0 27.5T Ω Ω Ω Ω 25T 22.5T

m 0.04 0 100 200 300 ( T (K) 20T 17.5T 15T 12.5T 0.02 10T 7.5T 5T 2.5T 1T 0.5T

Resistance 0.00 0T b

30 H H H 90 10 50 20

H (T)H 0 µ µ µ µ 10

26 28 30 32 34 36 38 40 T (K)

ρ(39 K). (b) Hc2(T) defined at 90% (H90), 50% (H50) and 10% (H10) resistance.

Figure 4.1 shows the magnetic Tc transition of both wire and bulk in a magnetic field of 2 mT. Both samples show a strong diamagnetic signal with little temperature dependence corresponding to superconducting volume fraction > 90%, indicating strong global screening currents crossing many high-angle grain boundaries. To characterize Hc2(T), the bulk resistance was measured by a 4-point method in magnetic fields up to 35 T. The resistivity measurements

can be found in figure 4.2. As seen in figure 4.3a, Hc2(0) is estimated above 90 T for K-doped 80 81 Ba122, well beyond the highest values obtained for Nb3Sn wires and MgB2 thin films and comparable to the K-doped Ba122 single crystal2 plotted for comparison. Figure 4.3b shows the 2 bulk and single crystal (H//ab) data normalized by their respective Tc (bulk: 37.4 K, crystal: 38.7 K). The excellent agreement with single crystal data also confirms that this bulk polycrystal 2 is of high quality since Hc2 of this compound has been shown to be very sensitive to doping.

100 1.0 H//ab 80 H//ab 0.8 Single H//c 60 crystal 0.6 Single

Bulk ( T/K ) (T/K c Crystal H T( ) 40 0.4 0 H/T µ µ µ µ 0 µ µ MgB µ µ 20 2 Bulk 0.2 Nb Sn 0 3 0.0 0 10 20 30 40 0.85 0.90 0.95 1.00

T ( K ) T/Tc

Figure 4.3 - Upper critical field as a function of temperature. (a) Hc2(T) defined at 90% resistance for the K-doped Ba122 bulk compared to an optimally doped single crystal from 2 22 81 reference, a Nb3Sn wire from reference, and a textured MgB2 thin film from reference with H applied parallel (closed symbols) and orthogonal (open symbols) to its surface. The dotted line 2 is a rescaled fit from reference to guide the eye. (b) Hc2 and temperature normalized by Tc to show close agreement between bulk polycrystal and single crystal with H//ab.

Figure 4.4 shows a transmission electron microscopy (TEM) image of the polycrystalline K- doped bulk. The diffraction contrast in figure 4.4a clearly shows that the average grain size is approximately 200 nm. The electron diffraction pattern from the selected area of Figure 4.4a indicates a randomly oriented polycrystalline structure containing many high-angle grain boundaries. The high resolution TEM observation confirms clean and well-connected grain boundaries in a randomly oriented polycrystalline bulk, as shown by a typical grain boundary in figure 4.4b. However, TEM images do reveal some porosity and secondary phase that obstruct

* FeAs 4000 (103)

2000 (116) (112) (213) (200)

(105) (008) (114) (110) (006) (100) (204) (004) ** * (202) Intensity (arb. unit) (arb. Intensity 0 * *

10 20 30 40 50 60

2θ (((CuΚαΚαΚα)Κα))) (((deg)

some current flow, indicating room for further process optimization. The impurity phase was identified by X-ray diffraction as FeAs and accounts for less than 3 % of the bulk volume by SEM image analysis (see figures 4.5 and 4.6). Figure 4.7 shows TEM and the electron diffraction pattern of Co-doped Ba122 material which has a microstructure that is similar to the K-doped

122. Magneto optical (MO) imaging was used to image the local field profile Bx produced by magnetization currents induced by magnetic fields of up to 120 mT applied perpendicular to the bulk sample’s surface. The MO images in figure 4.8 show a roof top pattern of magnetic flux density produced by bulk current flow over the entire ~3 mm-long sample, a length scale that is orders of magnitude larger than the ∼200 nm grain size seen in figure 4.8a. Figure 4.8a shows only a partial flux penetration due to strong induced currents caused by applying a magnetic field of 120 mT after zero-field-cooling (ZFC) the sample to 6 K. Figures 4.8b and 4.8c show uniform, fully-trapped, magnetic flux from applying a magnetic field of 120 mT and then field- cooling (FC) the sample from above Tc to 6 K and 32 K, respectively. The calculated current stream lines for the 32 K FC MO image in figure 4.8c are shown in figure 4.8d. Very uniform bulk current flow is still present above 30 K, a testament to the material’s electromagnetic

homogeneity and large superconducting volume fraction even close to Tc. MO images of a cross section of our K-doped wire (see supplementary figure 4.9) also show good electromagnetic homogeneity comparable to the bulk material. In contrast, MO images of other ferropnictide bulks show primarily granular currents indicating little or no bulk current flow4,11. Clearly, MO global indicates there is significant and well distributed Jc in our material.

Figure 4.9 - Magneto-optical images of a rectangular piece of K-doped Ba122 bulk material with magnetic fields applied perpendicular to plain of the sample (thickness = 0.7 mm). (a) Magneto-optical image of partial flux penetration after zero-field-cooling (ZFC) the sample to 6 K and applying a magnetic field of 120 mT. (b) Magneto-optical image of trapped magnetic flux in a sample field-cooled (FC) to 6 K in an external magnetic field of 120 mT. (c) Magneto- optical image of trapped magnetic flux in a sample FC to 32 K in an external magnetic field of 120 mT. (d) Current stream lines calculated for c that illustrate the uniform current distribution that circulates inside the bulk even near Tc.

magnetization transport Figure 4.10 shows Jc and Jc of the wire plotted as a function of applied magnetic 45,82 transport fields along with other round untextured Fe-based superconducting wires . Jc , measured for H perpendicular to the wire’s length, was calculated from the critical current (Ic) using the

area of the superconducting cross section of the wire (see figure 4.10 inset). Ic was determined

-1 magnetization using the electric field criterion Ec = 1 µVcm (See figure 4.11 for I-V curves). Jc was calculated from VSM measurements using the Bean model. The good agreement between the local transport and magnetization measurements indicates the intragrain Jc component makes only a global -2 small contribution to the magnetization. At self-field, a high Jc of over 0.12 MAcm is global global obtained, the highest reported Jc of any ferropnictide bulk or wire so far. Jc shows a weak field dependence and maintains a reasonably high value of 0.01 MAcm-2 at 12 T. This is over an order of magnitude higher than the best untextured wires and about 3 times higher than the textured Sr122 tape reported by Gao et al. at 10 T.73 Not only are these values high for global ferropnictide materials, but they approach the Jc values desired for applications.

4.4 – Discussion An SEM image of the cross section of the K-doped Ba122 wire is displayed in the inset to figure 4.5. The round wires made by PIT processing are advantageous since they can be made cheaply

106

105

) 4 2 10

K-doped Ba-122

3 (A/cm 10 c Co-doped Ba-122 J

102 Sm-1111

101 FeSe

0 5 10 15 Magnetic Field (T)

transport magnetization Figure 4.10 - Jc (symbols) and Jc (solid lines) as a function of applied magnetic field at 4.2 K for the K-doped wire compared to other round, untextured, Fe-based superconducting wires. Sm1111 wire is from reference45 and FeSe wire is from reference.82 Inset is an SEM image of the K-doped mono-core wire showing the round cross section with Ag and Cu sheaths.

a b 10 10 0T 0T 8 0.1T 8 0.1T 0.5T 1T 1T 2T 6 2T 6 3T V) V) µ µ µ µ µ µ 3T µ µ 5T 5T 4 4 9T 12T

Voltage ( Voltage 2 15T ( Voltage 2

0 0

10 100 1 10 100 Current (A) Current (A)

Figure 4.11 - I-V curves for the Ba122 wires at different fields. (a) K-doped Ba122 wire measured in fields up 15 T. (b) Co-doped Ba122 wire measured in fields up to 5 T. The Co-doped wire was made by the same PIT process used for the K-doped wire. Voltage response was measured at 4.2 K on a 4 cm pieces of wire with voltage taps approximately 1 cm apart. Measurements were made with increasing current.

and applied to traditional designs that depend on electromagnetically isotropic round wires. Since the superconductor is made ex situ and the final heat treatment is short and does not exceed 600 °C, sheath materials other than Ag may also be used without significantly degrading the conductor due to chemical reactions that occur with the sheath material at high temperatures. This may allow for the use of stronger, less expensive sheath materials. We have yet to explore in detail the effect of texturing, adding additional dopants, multi-core wires, or over-doping K, 44,73,83,84 all of which have been shown to increase Jc in ferropnictide wires or tapes .

global We propose that the unexpectedly high Jc arises through a combination of factors. First, our heat treatment occurs at a sufficiently low temperature that secondary phases do not wet grain boundaries, as earlier studies showed that such phases block current23. Second, high pressure synthesis results in nearly 100% dense material, which further contributes to good connectivity. Third, the fine grain size makes planar GBs very rare and the low γ value makes the vortex stiffness high. Thus, although essentially all vortices cross GBs, which may have a depressed superconducting order parameter, the GB vortex portion is short and can be anchored by the strong pinning of the superconducting segments lying in the grains. This situation has been studied for YBCO bicrystals with planar GBs by varying the angle between the B vector and the GB plane12,85. Only when the two are close and a significant length of vortex lies in the GB is the

GB Jc depressed below the intragrain Jc value. In our K-doped Ba122 bulks and wires, which have very small grains and thus a high density of non-planar GBs plus a small γ, we may expect that very little of the vortices actually lies in any GB. A final possibility is that the K-doped compound may have less depressed GB order parameters, perhaps due to a higher carrier density global induced by K segregation to the grain boundaries. The higher Jc of the K-doped wire compared to the Co-doped wire plotted in figure 4.10 suggests that particular compound-related factors may also be playing a role, whether due to differences in GB properties86 or to the role of hole (K), rather than electron (Co) doping in Ba122. Bicrystal experiments on the K-doped Ba122 will be valuable to explore the specific properties of planar grain boundaries that are possibly less weak linked than in other superconductors.

CHAPTER FIVE

DEPENDENCE OF K-DOPED BaFe2As2 SUPERCONDUCTING PROPERTIES ON SINTERING TEMPERATURE IN WIRES AND TAPES

Wires and rolled tapes of K-doped BaFe2As2 were produced by an ex-situ powder-in-tube processing route. The evolution of superconducting transition temperature (Tc) and remanent magnetic moment as a function of maximum applied fields was studied systematically for sintering temperatures between 550 ºC and 900 ºC. We find that while the onset Tc reaches a maximum > 37 K in our wires at a sintering temperature of 800 ºC, the maximum trapped remanent moment increases monotonically with decreasing heat treatment temperature to 550 ºC. global The global critical current density (Jc ) is found to be inversely proportional to grain radius (r), reaching over 100 kAcm-2 (5 K, 0 T) for wires and 180 kAcm-2 (5 K, 0 T) for tapes heat treated at 550 ºC. The 1/r dependence of Jc suggests that decreasing grain size further is a viable global route to improving Jc in K-doped BaFe2As2.

5.1 – Introduction

Wires of superconducting K-doped BaFe2As2 (Ba122) have now been produced with critical -2 -2 6 current density (Jc) exceeding 100 kAcm (0 T, 4.2 K) and 10 kAcm (10 T, 4.2 K). Jc (10 T, 4.2 K) can be increased close to an order of magnitude by mechanically inducing texture when rolling Ba122 wires into thin tapes,87 values very much in the realm of being interesting for applications. While much effort has been put into studying the dependence of superconducting properties on doping and various processing routes, systematic studies on the evolution of superconducting properties with heat treatment temperature for AE122 (AE = Ba or Sr) superconductors are lacking. Zhang et al. studied the effect of processing temperature on superconducting properties in Sr122 using a single-step in-situ powder-in-tube (PIT) reaction route at ambient pressure.88,89 They found decreasing FeAs impurity phases with increasing temperature up to 850 ºC along with a changing ratio of lattice parameters (a/c). They also global report that the global critical current density Jc increases monotonically with increasing heat treatment temperature (Theat treatment). Here we report a contradictory result obtained by an ex-situ

PIT processing route in which Jc increases with decreasing sintering temperature down to 550 ºC, benefiting substantially from a fine-grain microstructure analogous to Nb3Sn and MgB2.

5.2 – Experimental Details 42 Ba0.6K0.4Fe2As2 material was synthesized by a mechanochemical reaction path and then made into superconducting wire by the powder-in-tube technique described elsewhere.6 Ag/Cu tubes were used for samples heat treated below 800 ºC, and Ag/stainless steel tubes were used at higher temperatures to avoid the Ag/Cu eutectic. Tapes were made by rolling some of the wires which had an initial superconducting diameter of ~0.4 mm to a final core thickness of ~0.2 mm. Sealed pieces of the wire were then sintered under 192 MPa of pressure for 10 hours in a hot isostatic press at various temperatures. A magnetic property measurement system (Quantum design: MPMS-XL5s) was used to measure the magnetization as a function of increasing temperature after zero-field cooling (ZFC) the sample to 5 K and applying a magnetic field of 20 Oe. The same system was used to measure the remanent magnetization trapped in the sample as a function of increasing maximum applied fields as described in chapter 2 and by Müller et al.54 A scanning electron microscope (SEM) (Zeiss model 1540) was used to observe the microstructure of freshly cleaved wire cross sections.

5.3 – Results

Figure 5.1(a) and 5.1(c) show the remanent magnetization (MR) as a function of increasing maximum applied field (Hmax) after heat treatments at various sintering temperatures for the wires and tapes respectively. For these measurements, a maximum magnetic field is applied and

removed, the remanent magnetization (MR) is measured, and this is repeated in progressively increasing Hmax. Initially, the magnetic field is shielded by the superconductor via induced surface currents and the Meissner effect, but after high enough fields are applied, flux begins to penetrate into the sample where it becomes trapped, contributing to the measured MR. For a superconductor with poorly connected grains, flux penetrates preferentially at the grain boundaries.10,12,23,54 The resulting trapped magnetic flux results in global current flow global global throughout the bulk that produces a global remanent magnetization (MR ). MR saturates

when Happ is high enough for the magnetic field to completely penetrate the sample, putting it in the critical state. As the magnetic field is increased further, magnetic flux penetrates into the grains and becomes trapped which produces current loops on the scale of the grain size that sum

Figure 5.1 - Remanent magnetization at 0 T applied field as a function of maximum applied field for (a) wires and (c) tapes. Derivative of data in a and c is presented in (b) and (d) respectively.

local to produce a local remanent magnetization (MR ). The samples with very small grain size local global local have MR << MR , since MR scales with the size of the grains, so most of MR comes from the contribution of bulk current flow. In contrast, large-grain or poorly connected samples local global local can have MR > MR . Unlike MR , which saturates once flux fully penetrates the grains, global MR first increases and then decreases with increasing Hmax, hence the appearance of the shoulders in figures 5.1(a) and 5.1(b). This observed drop in MR can be attributed to the hysteretic behavior of the intergranular critical current density caused by trapped magnetic flux in the grains.54–56 Figures 5.1(b) and 5.1(d) show the derivative of the data in figure 5.1(a) and 5.1(c), respectively. For most samples, two distinct peaks are observed in figures 5.1(b) and

5.1(d) for each sample, corresponding to the field required for flux to penetrate the bulk (low field) and grains (high field). 10,12,23,54

Figures 5.2(a) and 5.2(b) show the magnetic moment as a function of temperature and heat

treatment. Any background moment above Tc has been subtracted from the data and the magnetization data has been normalized for clarity. Each sample had a volumetric susceptibility under -0.85, confirming bulk superconductivity within the samples. The samples sintered at 750 ºC and 800 ºC show the sharpest transitions at the highest temperatures. Figure 5.2(c) shows the superconducting transition temperature (Tc) as a function of heat treatment temperature.

Presented are three different definitions for Tc taken from the ZFC magnetization data. The onset 10 90 onset Tc (Tc ), Tc , and Tc are taken at 0.1%, 10%, and 90% of the normalized moment

respectively. Both wires and tapes show a similar evolution of Tc with heat treatment. Figure global 5.2(d) shows the self-field (SF) Jc calculated by the following approximation:

2 = -1 Where HP1 is the Hmax in Acm corresponding to the first peak on the plots in Figure 5.1(b) and 5.1(d), and a is the minor sample thickness in cm perpendicular to the applied field. This approximation neglects Jc anisotropy and the demagnetization factor. Similarly, assuming spherical grains, the local critical current density can be approximated by:

2.81 = where r is the grain radius. This calculation gives reasonable results for large grained samples (approximating local critical current densities on the order of those measured in single crystals)10 but fails for samples with r < λ since the length-scale of shielding currents is less than the size of the grains (see chapter 6). For our K-doped wires this calculation consistently gives values over 10 MAcm-2, which are uncharacteristically high. Therefore, it is more reasonable to estimate local local 54 Jc using the magnitude of MR at very high Hmax according to:

3 ( → ∞) =

local global where MR (Hmax → ∞) can be measured directly by breaking up the sample (to disrupt MR global without decreasing r). A less accurate approach that we use here is to subtract MR (Hmax → local ∞) ≈ MR(Hmax = 2Hp1) from MR(Hmax → ∞) to get MR (Hmax → ∞). This neglects the global hysteretic field dependency of Jc with Hmax, so it overestimates Jc and is only valid for local local samples with a considerable contribution of MR to MR. Using this approach, Jc was found to be between 2 and 6 MAcm-2, which are more reasonable values.

0.0 (c) (a) Wires -0.1 35 heat treatment -0.2 T 500 °C -0.3 550 °C 30 -0.4 600 °C -0.5 700 °C

750 °C (K) 25 c

-0.6 800 °C T Wires Tapes -0.7 900 °C T onset 20 c

Normalized moment -0.8 10% T c -0.9 90% T c -1.0 15 10 15 20 25 30 35 500 600 700 800 900 heat treatment Temperature (K) T (°°°C) (b) (d) 0.0 Tapes 200 5K, SF Grain radius (nm) -0.1 1500 200 70 heat treatment 180

) 100

-0.2 T -2 500 °C 160 Tape

-0.3 550 °C 140 50 ) (kAcm -0.4 600 °C -2 120 global c

750 °C J 0 -0.5 0.005 0.010 100 800 °C 1/Grain radius (1/nm) (kAcm -0.6 900 °C 80 Wire global -0.7 c 60 J

Normalized moment -0.8 40 -0.9 20 -1.0 0 10 15 20 25 30 35 500 600 700 800 900

heat treatment Temperature (K) T (°°°C)

Figure 5.3 shows SEM micrographs of each of the wires presented in figure 5.2(a). Each micrograph is labeled according to the heat treatment temperature. An optical micrograph is presented in the inset to the sample heat treated at 900 ºC showing impurities. The light contrast in the optical inset corresponds to Fe and the darker contrast corresponds to Ba or K. Samples heat treated below 700 ºC have sub-micron grain size.

900 °C 800 °C

Ba,K 10 μm 2 μm 2 μm 750 °C 700 C

2 μm 2 μm 600 °C 550 °C

1 μm 2 μm 1 μm 2 μm 500 °C

1 μm 2 μm

5.4 – Discussion The remanent magnetization data presented in figure 5.1 show a systematic decrease of global maximum MR with increasing heat treatment temperature. Both MR and Hp1 shifts to lower values as the sintering temperature is increased. While the sample size (a) can be controlled after local global heat treatment to remain constant; r, Jc and Jc can vary with sintering temperature.

Interestingly, Hp2 also shifts to lower values with increasing sintering temperature suggesting local that Jc must also be decreasing since r only increases with sintering temperature. One can assume that higher sintering temperatures are associated with less defects within the crystallites than samples sintered at lower temperatures, corresponding to a lower intrinsic pinning potential.

Tc increases with increasing heat treatment temperature up to 800 °C where it then decreases as shown in figure 5.2(c). The increasing Tc is not surprising, since the high amount of disorder in the samples sintered at low temperature, which can cause variations in the superconducting order parameter, is annealed out at higher temperature. However, at 900 °C Tc decreases again. This corresponds to an increase of Fe and Ba/K impurities, suggesting that As is being lost at higher temperatures. Arsenic deficiency after heat treatment has been reported90 and other groups commonly add extra As to the starting materials to compensate for similar reported 84,88,89,91,92 global losses. Figure 5.2(d) shows Jc calculated from the remanent magnetization data decreases with increasing sintering temperature. This is in contrast to what has been reported by 88,89 global Zheng et al. However, their Jc calculations are based on magnetization measurements, assuming current is flowing over the entire bulk sample. What we show here (as has been discussed elsewhere)10,12 is that there are two distinct scales of current flow contributing significantly to the magnetization of AE122 superconductors, one on the length scale of grains, and another on the length scale of the bulk. While it is difficult to determine the contribution of each in M-Happ loops, the systematic increase of Jc reported correlates mostly with an increase of grain size, suggesting that the magnetization being measured is mostly from intergranular currents that also scale with grain size. In addition, the addition of extra As and K could artificially lower properties at low annealing temperature since the resulting material still has excess As and K at low temperature, but the right stoichiometry at high temperature where these excess elements presumably diffuse out of the sample at a high rate. The application of external isostatic pressure in our case is another variable that allows full densification at lower temperature than solid state sintering at ambient pressure, where the only driving force for 55

densification is to minimize the surface free energy by capillary forces that are diffusion dependent.

global The tapes measured have a higher Jc overall, seen in figure 5.2(d), suggesting that rolling global assists with grain alignment that improves Jc similar to what has been reported by many other groups.73,92–94 The degree of alignment and microstructure was not studied for our tapes to compare with the wires. The anisotropy caused by rolling makes analysis of the magnetization data less straight forward than for the round wires with randomly oriented grains.

global The inset to figure 5.2(d) shows that the SF Jc scales linearly with 1/r, with the exception of the sample heat treated at 500 °C that had slightly smaller grain size than the sample heat treated heat at 550 °C but severely degraded Tc compared to all the other samples. This discrepancy at T treatment = 500 °C may be from low connectivity caused by incomplete sintering at such a low temperature. For the other samples, we suggest that the increase of grain boundary density is a route to produce high bulk transport currents in randomly oriented polycrystals of K-doped

Ba122. The pinning force density has been shown to be proportional to 1/r in MgB2 and 95,96 Nb3Sn, suggesting the relationship reported here may be a result of enhanced grain boundary pinning with decreased grain size. However, unlike MgB2 and Nb3Sn, Ba122 has weak linked GBs and very high intrinsic pinning within the grains. As a result, the mechanism may be different, where either the Josephson current density across GBs or the percolation of vortices is dependent on the GB density. This is discussed further in chapter 6. Assuming the Jc ~ 1/r relationship observed in figure 5.2 (d) continues to smaller grain sizes, a reduction of the grain global diameter of K-doped Ba122 to ~30 nm will improve Jc (SF, 4.2K) by at least a factor of five.

5.5 – Conclusions global In summary, the global critical current density (Jc ) was found to be inversely proportional to grain radius (r), reaching over 100 kAcm-2 (5 K, 0 T) for wires and 180 kAcm-2 (5 K, 0 T) for tapes with a grain radius around 80 nm. While attempts to further decrease the grain size by global lowering the sintering temperature to 500 °C resulted in material with degraded Tc and Jc , the

1/r dependence of Jc suggests that decreasing grain size further while maintaining bulk global homogeneity and granular connectivity may be a viable route to improving Jc in K-doped

BaFe2As2 round wires and tapes.

CHAPTER SIX

UNDERSTANDING WEAK LINKS IN Co AND K-DOPED BaFe2As2

The following chapter is an attempt at connecting certain electromagnetic phenomena that occur to the weakly linked grain boundaries in Ba122 materials. Vibrating sample magnetometer measurements were made at the Atominstitut at the Vienna University of Technology by Johannes Hecher working under Michael Eisterer. Their group is currently preparing a manuscript on theory of some of the same phenomenon discussed below, but their models will not be discussed here. Instead a more phenomenological explanation is proposed. Eric Hellstrom and David Larbalestier directed the research.

The electromagnetic response of Co- and K-doped BaFe2As2 superconductors with various grain sizes are studied to separate the contributions of inter- and intragranular induced currents. Intragranular magnetization can be suppressed with decreasing the grain size, revealing a large global hysteresis of Jc . For K-doped material with grain radius less than 100 nm, intergranular currents dominate the magnetization. We show evidence that either current or flux is percolating through the conductor by transport and magnetization measurements.

6.1 – Introduction As the primary source of defects in most materials, grain boundaries (GBs) play a large role in defining the macroscopic properties of a material.12,17,25,74,97 This is of particular importance when realizing applications for high temperature superconductors (HTS) that have very high intrinsic critical current densities that are suppressed by the presence of weakly linked GBs. Iron based superconductors (FBS) appear to be weakly linked, as is the case for cuprate HTS superconductors, though bicrystal experiments have shown critical currents in doped BaFe2As2 (Ba122) FBSs to be less sensitive to the misorientation of 001 tilt GBs.5,14 Bulk K-doped Ba122 materials with randomly oriented grains have been synthesized with significant global critical global -2 6,42 global current densities (Jc ) on the order of 10 kAcm . Even in these high Jc samples, observation of oxygen impurities and stoichiometry variation across GBs on the order of a couple of coherence lengths (ξ) suggests there may be still be a significant extrinsic component to current blocking.63,90

Routes to synthesize long length superconducting wires out of weakly linked material typically involve texturing bulk polycrystals to align the crystallites favorably, which results in a flat wire or tape that limits or complicates their application. Ba122 conductors have been grown as a thin 19,98,99 film on textured substrates, a technology developed for the growth of YBa2Cu3O7-x

(YBCO), though the low Tc and Jc compared to YBCO doesn’t justify this costly approach to conductor development. The powder-in-tube technique has the potential to produce long-length global conductors cheaply, and efforts to improve Jc of Ba122 and Sr122 conductors by rolling or global pressing to align grains have now managed to produce Jc around the level desired for high field applications (> 0.1 MAcm-2 at > 4.2 K and >10 T),15,92 and by a rolling technique that is 74,100 actively employed for the production of Bi2Sr2Ca2Cu3O10+x (Bi-2223) tapes. While inducing texture for 2223 by rolling to make a viable technology is one proven path to applications, it is less than ideal, and the study of the influence of GBs on electromagnetic behavior in general is justified as it may allude to new options for overcoming weak links.

6.2 – Experimental Details

(Ba0.6K0.4)Fe2As2 and Ba(Fe0.92Co0.08)2As2 materials were synthesized by a mechanochemical reaction path.42 After an initial heat treatment, the material was milled and either sintered at various temperatures by the same route as referenced or made into superconducting wire by the powder-in-tube technique described elsewhere.6 Sealed pieces of the wire were sintered under 192 MPa of pressure for 10 hours in a hot isostatic press at 800 °C to produce a coarse-grain microstructure (average grain radius ~ 2-4 μm) and 600 °C to produce a fine-grain microstructure (average grain radius ~ 75-100 nm). A magnetic property measurement system (Quantum design: MPMS-XL5s) was used to measure the magnetization as a function of increasing temperature after zero-field cooling (ZFC) the sample to 5 K and applying a magnetic field of 20 Oe. The same system was used to measure the remanent magnetization trapped in the sample as a function of increasing maximum applied fields as described in section 2.4.1.2. A 15

T Oxford magnet system was used to measure transport Jc as a function of applied fields, and a 5 T vibrating sample magnetometer at the Vienna University of Technology was used to measure the magnetization as a function of applied field.

6.3 – Results and Discussion 6.3.1 – The chemistry of grain boundaries GBs are high energy interfaces between crystallites. With 5 degrees of macroscopic freedom between the millions of crystallites within a bulk, the properties of individual GBs can vary enormously. The bicrystal experiments discussed earlier are on idealized tilt boundaries and represent a tiny fraction of all the interfaces possible between two grains. The high energy of GBs provides a driving force for the segregation of atoms, vacancies, and impurities. While GBs are technically an interface, the high density of periodic charged dislocation cores is associated with a strain field that can extend several angstroms into the crystallites.17,25,101,102 The disorder associated with GBs is enough to suppress the superconducting order parameter in dislocation channels and sometimes kill superconductivity all-together, hence the term weak-links.17,102,103

This can be advantageous, as is the case for superconductors like Nb3Sn and MgB2 that rely on GB pinning, or detrimental as is the case for YBCO. Cuprates and FBS are particularly sensitive to GB interfaces because of their anisotropy and short coherence lengths. In addition, their superconductivity arises due to doping. In the case of the cuprates, oxygen is the dopant. Oxygen is extremely mobile allowing the dopant level to be altered easily by annealing in oxygen environments. While advantageous for tuning the properties of single crystals, this high mobility makes controlling the oxygen content across grain boundaries difficult if there is a changing chemical potential across the GBs that would attract oxygen vacancies or interstitials.

global While Co-doped BaFe2As2 polycrystals seem to have very low Jc in polycrystals, K-doped

BaFe2As2 have over an order of magnitude higher values. One possible explanation lies in the sensitivity to chemical off-stoichiometry within the proximity of GBs. Figure 6.1(a) shows a max comparison of normalized superconducting transition temperature (Tc/Tc ) as a function of doping. Co-doped Ba122 and YBCO have a much narrower superconducting phase space, indicating that small deviations from ideal stoichiometry will result in strongly suppressed superconductivity, while K-doped Ba122 has a relatively broad phase space. This is one possible

explanation for the sharper Tc transitions (see figure 6.1(b)) in susceptibility measurements and global higher Jc for K-doped Ba122 compared to Co-doped Ba122 and YBCO polycrystals. The samples in figure 6.2(b) were all polycrystalline wires with similar grain size between 1 and 5 μm.

6.3.2 –Intergranular vs intragranular magnetization

As described in section 2.4.1.2 and chapter 5, trapped remanent field measurements (MR) at Happ

= 0 made after incrementally applied maximum fields (Hmax) allow us to separate multiple scales of current flow (here we are interested in global, intergranular, current and in local, intragranular, current). The critical current density of each scale of current flow at self-field can then be calculated using the Bean model and the applied fields at which flux fully penetrates the volumes in which current is induced. This is only true as long as:

1) The global critical state is reached at a lower applied field than the local critical state, since bulk shielding currents would otherwise shield flux from penetrating into the grains.

2) The grain size is larger than 2λ, since screening currents induced on the grain’s length scale encompass the entire grain by the time the vortices have reached the surface of the grains. The vortex dynamics in this case are extremely complicated. One may expect an absence of driving force for flux to penetrate into the grains until the vortex lattice spacing approximates half the grain diameter (i.e. one vortex per grain).

Figures 6.2(a) and 6.2(b) show MR as a function of Hmax for Co- and K-doped Ba122 bulks after sintering at different temperatures. The insets for figures 6.2(a) and 6.2(b) show the derivative of the remanent magnetization for the Co- and K-doped Ba122, respectively. The peaks in the derivative remanent magnetization can be correlated to different scales of current flow.10,12,23,54

The first peak (at lower field) Hp1 for each sample corresponds to the applied field at which flux penetrates into the bulk superconductor, and the second peak (at higher field) Hp2 indicates the applied field at which flux penetrates into the grains of the superconductor. The location of the global local peaks can be used to estimate Jc and Jc as long as the Bean model is valid, since Hp ~ Jc x (size of current loops). For the K-doped material, it takes much higher fields to penetrate the global sample, corresponding to higher Jc . In addition, for the K-doped sample sintered at 600 °C, global local global the small grain size and high Jc results in MR << MR , so most of MR comes from the bulk current flow. This is demonstrated by figure 6.3 in which the bulk sample was ground in a mortar and pestle and then sieved to produce powder samples with different sizes of

conglomerates. This destroys the bulk connectivity and its contribution to MR. Despite having similar sample mass and being normalized by mass, only a very small signal is shown for the

local local 20/25 μm powder size confirming that MR is not contributing significantly to MR. MR 54 (Hmax → ∞) can be calculated by the following formula:

= 3 local 2 local where r is the grain radius. For Jc = 1 MA/cm and r = 80 nm, we get an expected MR = 0.267 emu/cm3 that matches well with what is measured in figure 6.3 after grinding. However, local for samples with poor grain to grain connectivity or large grain size, MR can contribute local significantly to or dominate MR. Unlike MR , which saturates once flux fully penetrates the global grains, MR first increases and then decreases with increasing Hmax, hence the appearance of global the shoulders in MR data where most of the remanent moment comes from MR . This observed drop in MR is attributed to the hysteretic behavior of the intergranular critical current density caused by trapped magnetic flux in the grains.54–56

global Figure 6.4 shows Jc vs. the sintering temperature calculated from Hp1 for K-doped wires magnetization along with the global magnetization critical current density Jc calculated using the Bean model from magnetization measurements taken in a vibrating sample magnetometer (VSM). magnetization global The Jc assumes the entire magnetic moment (m) is due to Jc , which is only a good 62

approximation as long as mlocal << mglobal since they both contribute to m. This is the case for the samples annealed at low temperatures, but at Tanneal > 600 °C, grain growth and decreased connectivity likely results in mlocal > mglobal, which is why the magnetization measurements global overestimate Jc as the sintering temperature is increased.

6.3.3 – Percolation of current paths To determine if current can be applied parallel to the flux lines in a force-free configuration,107 transport we measured the transport critical current (Jc ) as a function of applied field (μ0Happ) in both increasing and decreasing fields applied perpendicular (H┴) and parallel (H//) to the direction of applied current. For H┴, a maximum Lorentz force can be expected since the field is perpendicular to the direction of current. However, for H//, we expect a significant increase of transport Jc since there is little Lorentz force to drive the movement of magnetic flux that results in dissipation. However, if the current path is not parallel to the flux lines, as is the case for percolative current or flux flow, the arrangement will not satisfy the force-free condition (B x J =

0). An increase of Jc proportional to 1/sin(θ) where θ is the angle between I and Happ has been 108 transport reported for Nb3Sn. The Jc (μ0Happ) data in figure 6.5 shows little difference whether H is applied perpendicular or parallel to current, suggesting current or flux lines have to percolate

around defects, the most likely source of which are weak links at grain boundaries. Due to two distinct scales of current flow observed in the previous section and the large hysteresis of transport current seen in figure 6.5 and discussed in the following section, chances are flux lines percolate along grain-boundaries where they bend with respect to the applied field, though there’s likely a contribution of both percolating flux and percolating current.

6.3.4 – Irreversible intergranular critical current density in applied fields The irreversible critical current dependency on increasing and decreasing applied fields as shown in figure 6.5 by transport and figure 6.6 by magnetization is a very well documented phenomenon often observed in weak-linked granular superconductors.55,100,109–114 For decreasing fields, the critical current (Ic) was greater than for increasing ones; a phenomenon most authors attribute to the sensitivity of critical current to local intergranular fields as modeled by Evetts and Glowacki.111 This hysteresis has also been modeled in terms of the demagnetization effects of the grains on the grain boundaries which carry some field dependent Josephson current.110 These

models fit data quite well for Bi-2223 tapes at low applied fields, matching both the hysteresis in

Jc and the anomalous peak in the intergranular magnetic moment on decreasing positive applied field (see inset to figure 6.6). However, this explanation fails to explain the presence of large hysteresis in very high applied fields. D’yachenko et al. developed a model to describe such behavior in terms of the current density at the surface of a grain being the sum of the Meissner shielding current density and the intragranular critical current density of the grain, which has a different sign on increasing and decreasing field, causing the transport current to become irreversible.55,56 This model can account for an irreversible transport current up to very high fields, but fails to explain quantitatively the maximum applied field dependence of the anomalous peak of the intergranular magnetic moment on decreasing field as displayed in the inset of figure 6.6. Both models were developed for superconductors with grain size (r) >> λ and there is not an adequate model for granular superconductors with very small grain size (r < λ), though we have collaborators actively modeling these materials.

We suggest preferential flux penetration and movement at grain boundaries is a percolative process, and an increase of the density of percolative paths for flux motion prevents flux lines from piling up densely at GBs while they are shielded by the surface currents of individual grains. For r < λ, surface currents penetrate the entire grains as soon as flux penetrates the GB 65

network. One implication is that the field required for vortices to penetrate the grains, where

they can be strongly pinned, is much higher than Hc1 since the driving force for flux penetration into the grains is not present until the flux density exceeds the GB density. Assuming a hexagonal GB lattice, a GB critical matching field (HGB) can be defined:

= 1.075 Φ

Where Φ0 is the magnetic flux quantum and:

= 3

For r = 80 nm, this corresponds to a HGB of over 3000 Oe that needs to be applied before the flux global density exceeds the GB density. This is around the Hmax where MR and m(0 T) reach their maximum value before decreasing with further increasing Hmax in figures 6.3 and 6.6 respectively. Once this applied field is exceeded, vortices enter the grains where they can be strongly pinned. When the field is decreased, the flux lines that are weakly pinned within the GB network exit first. This lowers the flux density at the GBs for a given applied field in decreasing field compared to in increasing field, so there are fewer weakly-pinned vortices at the GBs susceptible to motion due to the Lorentz force in addition to less blockage of current across GBs from the non-superconducting vortex cores. These effects may contribute to the observed global hysteresis of Jc on decreasing field after large applied fields.

6.4 – Conclusions In summary, we showed how weak-links affect the electromagnetic response of Co- and K-

doped BaFe2As2 superconductors. The intrinsic contribution to magnetization measurements can global be suppressed with decreasing the grain size, revealing a large hysteresis of Jc . This hysteresis is similar to that observed in Bi-2223 and YBCO polycrystalline conductors, but is unique in that samples with very fine grain size require large applied fields to trap flux within the grains. Transport Jc does not vary significantly whether magnetic field is applied parallel or perpendicular to the direction of current, indicating that either current or flux is percolating through the conductor.

CHAPTER SEVEN

EVIDENCE FOR COMPOSITION VARIATION AND IMPURITY SEGREGATION AT GRAIN BOUNDARIES IN HIGH CURRENT DENSITY POLYCRYSTALLINE K- AND Co-DOPED BaFe2As2 SUPERCONDUCTORS

To study the chemistry of grain boundaries in Ba122 superconductors, atom probe tomography was performed at the Northwestern University Center for Atom Probe Tomography by Dr. Yoon-Jun Kim working under Professor David Seidman. This work was published in Applied Physics Letters.90 As second author, I took the lead for the FSU contribution to the experiment, synthesizing samples, performing electromagnetic characterization, and contributing significantly to the preparation of the chapter that follows. Eric Hellstrom and David Larbalestier directed FSU’s contribution to the research and all authors discussed the results and commented on the manuscript.

Some polycrystalline forms of the K- and Co-doped BaFe2As2 and SrFe2As2 superconductors

now have a critical current density (Jc) within a factor of ~5 of that required for real applications, even though it is known that some grain boundaries (GBs) block current, thus raising the question of whether this blocking is intrinsic or extrinsically limited by artefacts amenable to improvement by better processing. Herein we utilize atom-probe tomography (APT) to study the grain and GB composition in high Jc K- and Co-doped BaFe2As2 polycrystals. We find that all GBs studied show significant compositional variations on the scale of a few coherence lengths (ξ), as well as strong segregation of oxygen impurities, which we believe are largely introduced in the starting materials. Importantly, these findings demonstrate that APT enables quantitative analysis of the highest Jc K-doped BaFe2As2 samples, where analytical transmission electron microscopy (TEM) fails because of the great reactivity of thin TEM samples. The observations of major chemical perturbations at GBs make us cautiously optimistic that there is a large extrinsic component to the GB current blocking, which will be ameliorated by better processing, for which APT will likely be a crucial instrument.

7.1 – Introduction Iron pnictide superconductors form a class of layered Fe-As compounds that exhibit

9 superconductivity with a transition temperature, Tc, up to 56 K. Numerous studies have been performed to elucidate their structure-property relationships and it is now quite clear that the superconducting properties of these low carrier density metals are strongly affected by composition,12 compositional inhomogeneities,2,98 impurities,115 and grain boundaries.5,12,14 It

5,12,14 was earlier found that the critical current density (Jc) of Co-doped BaFe2As2 (Ba122) [001] tilt thin-film bicrystals decreased quasi-exponentially as a function of grain boundary (GB)

17 misorientation angle, much as in YBa2Cu3O7-x (YBCO), though at a less steep rate. In an attempt to explore the capabilities of polycrystalline forms for applications, several groups have 6,11,43,50,116 explored the Jc values of textured and untextured tapes, bulks and round wires. An

6 4 2 important breakthrough was made by Weiss et al. who demonstrated that Jc > 10 A/cm could be obtained in fine-grain (<100 nm) K-doped Ba122 bulk and wire forms without a global 23 texture, where the wetting Fe-As phase responsible for low Jc in so many bulk forms was avoided by reaction at no more than 600°C. Subsequent studies of rolled or pressed tapes of K-

15,92 doped Sr122 has raised Jc by several times, at least in the most favorable direction, suggesting that texture is valuable and some intrinsic super-current blocking occurs in randomly oriented GBs, not just [001] tilt GBs. There may, however, be only limited interest in a tape-form conductor given the competition from higher Tc tape-form YBCO coated conductors (Tc ~90 K) and tapes of (Bi,Pb)2Sr2Ca2Cu3O10 (Bi2223) (Tc ~100 K). As a round wire, however, there are many attractions of a 38 K superconductor like K-doped Ba122 with almost no electronic anisotropy (Hc2 anisotropy ~1.1) and Hc2(0) ~ 90 T, provided that the in-field current density can be raised to ~5 x 104 A cm-2 in the 10-30 T range at 4.2 K.22,117 It is therefore very important to understand if the properties of present Ba122 polycrystals are limited by intrinsic GB properties or by extrinsic factors, such as impurity segregation or chemical composition variations that take GB compositions far away from the bulk compositions. Transmission electron microscopy

(TEM) has revealed oxide phase formation at the GBs in a polycrystalline (Sr0.6K0.4)Fe2As2 superconductor,84 but slight compositional changes and segregation of lighter atoms on a smaller length scale can be difficult to detect and quantify by analytical TEM, especially due to the rapid degradation of electron-transparent samples, which has prevented accurate quantitative analyses

of GBs of the higher Jc K-doped 122 samples. 68

In this letter, we examine GB compositions of K- and Co-doped polycrystalline Ba122

superconductors, (Ba0.6K0.4)Fe2As2, (Ba0.4K0.6)Fe2As2, and Ba(Fe0.92Co0.08)2As2, employing atom-probe tomography (APT). APT characterization produces a three-dimensional (3-D) reconstruction of the lattice on an atom-by-atom basis, which permits an accurate chemical concentration measurement of all the elements in the periodic table with essentially the same detection efficiency. The 3-D atomic scale reconstruction of a sample is obtained by combining the times-of-flights (TOFs) and the x-, y-, and z-positional data of all the field-evaporated atoms in an analyzed volume with sub-nanoscale spatial resolution. Additionally, TOF mass spectra, based on mass-to-charge state (m/n) ratios, are utilized to obtain quantitative chemical analyses of the 3-D reconstructed sample.

7.2 – Experimental Details The samples were all prepared from pure elements (Alfa Aesar, Ba, Fe, K, As, and Co with 99%, 99.5%, 98%, 99%, and 99.8% purity, respectively). Although these all had high purity and were

handled in a glove box with <1 ppm O2, in fact they do introduce a measurable oxygen contamination, as described later. The elements were mixed to obtain polycrystals of nominal

compositions (Ba0.6K0.4)Fe2As2, (Ba0.4K0.6)Fe2As2, and Ba(Fe0.92Co0.08)2As2 and then ball milled for 1 hour. After ball milling they were wrapped with Nb foil and placed in stainless-steel ampoules that were evacuated, welded shut, compressed into pellets in a cold isostatic press at 275 MPa, and then heat treated in a hot isostatic press (HIP) under 192 MPa of Ar at 600oC for 20 hours, a temperature below that needed to form the GB-wetting FeAs phase. To obtain a more homogeneous chemical composition, the material was remilled and heat treated again at 600oC for 10 hours in the HIP. For the Ba(Fe0.92Co0.08)2As2 wires, the bulk material was milled for a third time and packed into a Ag tube (6.35 mm outer diameter, 4.35 mm inner diameter), whose ends were plugged, swaged and welded shut. The tube was then groove rolled, followed by drawing to a 0.8-mm-outer-diameter wire. Pieces of the Ag-clad wire were sealed in Cu tubing (1.57-mm-outer-diameter, 0.86 mm-inner-diameter) under vacuum by welding the ends shut. The Cu tubing was then groove rolled to an outer diameter of ~1.35 mm. The Cu/Ag-clad wires were then compressed in a cold isostatic press under 2 GPa pressure and further heat treated at 600oC for 10 hours in the HIP.

After the superconducting characterization described below, each sample was then fabricated into needle-shaped specimens with a microtip radius of <20 nm using a standard lift-out technique, utilizing a dual-beam focused-ion beam (FIB) microscope (FEI Helios Nanolab).118– 121 Microtip samples were immediately inserted in the APT’s ultrahigh vacuum chamber to minimize exposure to air, though some oxygen was detected on the surface of the microtip, which was removed by field-evaporation. A laser-assisted local-electrode APT (Cameca LEAP 4000XSi) was utilized, employing an ultraviolet picosecond laser with a wavelength of 355 nm, a pulse repetition rate of 250 kHz, and an energy per pulse of 20 pJ, resulting in an evaporation rate of 0.005 to 0.02 ion per pulse. The 3-D image reconstruction and chemical analyses were performed using the software package IVAS 3.6.6.

(a) (b)

5 2 mT ZFC 10

0.0

4 8% ) 10 40% K-doped -2 Co-doped

vol χ χ χ χ -0.5 60% 60% K-doped (Acm

K-doped c J 103

40% 8% Co-doped -1.0 K-doped

5 10 15 20 25 30 35 0 5 10 Temperature (K) Applied Magnetic Field (T)

7.3 – Results and Discussion Figure 7.1 displays the superconducting transitions as a function of increasing temperature after zero-field-cooling (ZFC) the samples to 5 K and then applying a magnetic field of 2 mT. All three samples exhibit sharp diamagnetic transitions indicative of macroscopically uniform and strong superconducting coupling across the whole sample, thus indicating that many, if not all,

GBs are carrying current. Figure 1b demonstrates that the magnetization Jc (calculated using the Bean model and the sample dimensions) in the K-doped samples approaches 105 A cm-2 at self-

field and 4.2 K, values very similar to those of previously studied samples of bulk or wire 6 specimens of (Ba0.6K0.4)Fe2As2 measured in transport. The Jc of the over-doped

(Ba0.4K0.6)Fe2As2 sample is similar to the optimally doped (Ba0.6K0.4)Fe2As2 sample, even though

it has a lower Tc of 28 K. For completeness we note that about a twice higher in-field Jc in

untextured bulks of (Ba0.6K0.4)Fe2As2 has been obtained by us in the best samples made by this route.42 These differences may be due to variations either of the vortex pinning strength or variations in the GB connectivity. Without accurate knowledge of the structure and composition of the GBs, supplied by experiments of the type reported herein, it is impossible to decide between these two possibilities.

Table 7.1 - Nominal and APT measured compositions of bulk Ba122 samples (at.%)

Ba K Fe Co As O (Ba0.6K0.4)Fe2As2 Nominal 12.00 8.00 40.00 - 40.00 0 Measured 13.37 9.03 42.62 - 33.67 1.32 ± 0.010 ± 0.008 ± 0.020 ± 0.017 ± 0.003 (Ba0.4K0.6)Fe2As2 Nominal 8.00 12.00 40.00 - 40.00 0 Measured 10.31 10.45 43.32 - 31.96 3.96 ± 0.020 ± 0.020 ± 0.047 ± 0.039 ± 0.012 Ba(Fe0.92Co0.08)2As2 Nominal 20.00 - 36.80 3.20 40.00 0 Measured 22.08 - 40.12 3.11 33.41 1.28 ± 0.011 ± 0.015 ± 0.004 ± 0.014 ± 0.002

Table 7.2 - Nominal and APT measured compositions inside crystallites (grains) within Ba122 samples (at.%)

Ba K Fe Co As O (Ba0.6K0.4)Fe2As2 Nominal 12.00 8.00 40.00 - 40.00 0 Measured 12.70 9.10 42.52 - 34.92 0.77 ± 0.28 ± 0.46 ± 0.58 ± 0.50 ± 0.09 (Ba0.4K0.6)Fe2As2 Nominal 8.00 12.00 40.00 - 40.00 0 Measured 8.47 11.84 44.32 - 34.26 1.10 ± 0.47 ± 0.74 ± 1.05 ± 0.94 ± 0.21 Ba(Fe0.92Co0.08)2As2 Nominal 20.00 - 36.80 3.20 40.00 0 Measured 22.02 - 40.58 3.14 33.49 0.77 ± 0.64 ± 0.73 ±0.22 ± 0.80 ± 0.21

The bulk chemical-compositions of each sample measured using APT are summarized in Table 7.1. These compositions are based on total ion counts of 12,489,403, 2,253,152, and 20,281,659

for (Ba0.6K0.4)Fe2As2, (Ba0.4K0.6)Fe2As2, Ba(Fe0.92Co0.08)2As2, respectively. All three samples are

globally Fe-rich and As-poor. For K-doped Ba122, the concentration of K is 1.03 at.% higher

and 1.55 at.% lower than the nominal compositions of (Ba0.6K0.4)Fe2As2 and (Ba0.4K0.6)Fe2As2, respectively. Similarly, Co-doped Ba122 contains 2.08 at.% higher Ba, 3.32 at.% higher Fe, and 6.59 at.% lower As concentrations with respect to the nominal composition of

Ba(Fe0.92Co0.08)2As2. The measured Co composition was 3.11 at.%, close to its nominal composition. Ba compositions were both under and over the nominal value. A significant O contamination was, however, also observed in both K samples and it was more marked in the over-doped sample compared to the optimally doped sample (3.96 vs. 1.32 at.% of oxygen). Very few detailed analyses of polycrystalline 122 samples have, however, been made, so that it is unclear whether these are typical or atypical of chemical composition variations in carefully processed samples. As noted, the strong shielding exhibited in the Tc plots suggests rather good superconducting uniformity, even given these global deviations from the expected compositions.

The Ba(Fe0.92Co0.08)2As2 sample contained 1.28 at.% O.

Figure 7.2 shows 3-D reconstructed images of the three polycrystalline Ba122 compositions that clearly indicate that O atoms (blue) can segregate strongly to the GBs of all 3 samples. For clarity only Ba (orange) and no K, Fe and As atoms are displayed in the 3-D reconstructed images. All GBs lie perpendicular to the images displayed in figure 7.2. In figure 7.2(a), the

optimally K-doped sample with the highest Jc, (Ba0.6K0.4)Fe2As2, exhibits the least oxidation, the majority of visible GBs being clean. Where oxygen is present, it appears densely on GBs. The over-doped K sample contains three GBs (GB1, GB2, and GB3 in figure 7.2(b)), which are much more fully oxidized, consistent with the higher bulk O concentration indicated in table 7.1. Contamination is even more complete in the more extensive GB network of the Co-doped

Ba(Fe0.92Co0.08)2As2 wire; the most clean GBs are indicated as GB1 and GB2 in figure 7.3(c). Note, however, that each sample has uncontaminated regions of GBs, across which we speculate super-current transport may be more favored.

Line scans across the GBs (figure 7.3) demonstrate clearly broader changes in the chemical compositions of GBs with respect to the grains themselves. In general, GBs are depleted in both Ba and Fe, and enriched in O. Based on the mass spectra obtained (data not shown), a peak appears at m/n = 91, which is suggestive of AsO as the GB oxide. The K-doped Ba122 samples, in particular, exhibit depletion of the K dopant at the GBs, figures 7.3(a) and 7.3(b). GB1 of the

(Ba0.6K0.4)Fe2As2 sample presented in figure 7.2(a) displays local minimum concentrations of

11.84 at.% Ba and 6.21 at.% K, while GB1 of the (Ba0.4K0.6)Fe2As2 specimen displayed in figure 7.2(b) has local minimum concentrations of 6.61 at.% Ba and 4.16 at.% K. Additionally, this sample exhibits local maximum concentrations of 38.15 at.% Fe and 9.34 at.% O at the GB, which may also be attributed to AsO. The concentration profile of the Co dopant in

Ba(Fe0.92Co0.08)2As2 is unchanged across the GB: figure 7.3(c) and 7.3(d) taken from GB1 and GB4 shown in Fig 2c. AsO is the only oxide type found in the K-doped Ba122 samples, whereas the Co-doped sample contains both AsO (figure 7.3(c)) and a Ba-rich oxide (figure 7.3(d)). Such oxide formations are correlated with local minimum or maximum concentration profiles across GBs. Figure 7.3(c) displays local maximum concentrations of 41.92 at.% As and 3.51 at.% O and a local minimum concentration of 14.91 at.% Ba. While figure 7.3(d) indicates local maximum concentrations of Ba and O of 26.10 and 7.90 at.%, respectively, which corresponds to a local minimum Fe concentration of 27.20 at. %.

Segregation of O atoms along a GB is strongly correlated with the formation of As-oxide and Ba-oxide, which results in significant compositional changes of the neighboring grains of the K-

and Co-doped Ba122 polycrystals. Interestingly, this does not degrade the susceptibility plots

(figure 7.1(a)), which display sharp superconducting transitions, though Tc for the samples was lower than for the best single crystals. Table 7.2 lists compositions measured inside the grains. Since O segregates at the GBs, the grains contain less O than the average bulk compositions of Table 7.1. The Ba and K (or Co) concentrations are higher and As is approximately 15% lower than the nominal (as weighed out) compositions of all three samples.

The meaning of all of these results is not yet clear, but one very important observation is that oxygen impurities concentrate at the GBs, suggesting that the critical current density properties have an extrinsic and not just an intrinsic component. Although O can be a beneficial dopant in the 1111 compounds where it, like F, dopes carriers into the Fe-As layers; it is far from certain that O is beneficial for the 122 compounds studied herein, especially as it forms As- and Ba-

oxides. Based on the sharp superconducting transitions and the rather high Jc values (about a factor of 10 less than single crystals and about 100 times less than Co-doped 122 thin films), we know that there is considerable GB transport in the superconducting state, much more than

122 occurs in (clean) untextured polycrystals of cuprates. Thus, the problem of how to increase Jc of the polycrystals seems to be a mixed intrinsic/extrinsic one. What is clear is that the APT can reveal this segregation.

If practical applications are to be achieved in Fe-based superconductors, they need to fill a niche where current superconducting technology is lacking. The high Hc2, of ~90 T, the intermediate Tc

of 35-38 K, and most importantly, the possibility of developing high Jc in round untextured wires by traditional industrial processes, may bring concomitant advantages in terms of price and production scalability for fields of greater than 10 T and temperatures above 20 K, where the only present alternatives are tape-form conductors. The presence of GB oxygen and the significant GB compositional variations on length scales of about 10 nm or 2-3 coherence lengths (ξ) sets up conditions for blocking of current at GBs. Yet actually the important result 5 4 -2 from a superconducting point-of-view is that Jc values of 10 -10 A cm can be obtained, even with these high GB densities in fine-grained samples. Clearly some GBs are effectively transmitting super-currents and perhaps it is the less contaminated GB regions that are indicated by the present APT study. We also note that the GB segregated regions observed herein are much thinner than those reported by Wang et al.84 Additionally, for the record we note that an

APT study on the conventional, s-wave superconductor Nb3Sn display moderate impurity

segregation of Cu to Nb3Sn GBs, which is not believed to harm the super-current transport properties.123 For these and other reasons, we believe that additional APT studies of GB chemistry and impurity segregation in combination with electromagnetic characterization and modeling, will be useful to better understand the fundamental properties of weak links in the Fe- based superconductor systems and to disentangle intrinsic from extrinsic properties.

7.4 – Conclusions In summary, overall compositions and grain boundary (GB) segregation of oxygen atoms in

three compositions of the pnictide superconductors, (Ba0.6K0.4)Fe2As2, (Ba0.4K0.6)Fe2As2, and

Ba(Fe0.92Co0.08)2As2, have been studied by local-electrode atom-probe (LEAP) tomography. We find significant variations of composition between grains and GBs and significant GB segregation of oxygen. Segregated oxygen atoms tend to react to form local As-oxide and Ba-

oxide on a width of approximately 5-10 nm in these high Jc samples, suggesting that there is still a large extrinsic component to the current blocking that may be eliminated by cleaner processing. APT provides the means to analyze even slight compositional variations in 3-D space with sub- nm to nm-scale resolution, making it an essential instrument for developing bulk superconductors with GBs that are free of these extrinsic current blockers.

CHAPTER EIGHT

DEMONSTRATION OF AN IRON-PNICTIDE BULK SUPERCONDUCTING MAGNET TRAPPING OVER 1T

In the following chapter, we utilized our best K-doped BaFe2As2 material to make bulk magnets. Magneto-optical imaging was done by Anatolii Polyanskii and the rest of the electromagnetic characterization was carried out at the University of Tokyo by Akiyasu Yamamoto. Ross Richardson did the hardness measurements. Eric Hellstrom directed the research and David Larbalestier provided guidance in preparation of the manuscript. The following chapter is in preparation for publication.

A trapped field of over 1 T (5 K) and 0.5 T (20 K) has been measured between a stack of magnetized cylinders of bulk polycrystalline Ba0.6K0.4Fe2As2 superconductors measuring 10 mm in diameter and 18 mm in combined thickness. The time dependence of the trapped field showed a low magnetic creep rate (~3% after 24 hours at 5 K), and magneto-optical imaging revealed a trapped field distribution corresponding to uniform macroscopic current loops circulating through the sample. The superconductors were manufactured by hot isostatic pressing pre- reacted powders using the scalable powder-in-tube technique. Vickers hardness indentations indicate that the bulk material has high hardness ~3.5 GPa and a fracture toughness ~2.35 MPa 0.5 m . Given that the sample diameter is relatively small with Ba0.6K0.4Fe2As2 showing large irreversibility field (> 90 T) and small decay of Jc in high fields, larger bulks are expected to trap much higher fields, in excess of 10 T.

8.1 – Introduction Since their discovery in 2008,1 a tremendous research effort has been made to synthesize and study Iron-based superconductors. Much of this effort has been driven by reports of properties that are very appealing for applications including low anisotropy around 1-2,66 high upper critical

2,124 -2 fields (Hc2) in excess of 90 T, and intrinsic critical current densities above 1 MAcm (0 T, 4.2 K).4 Unfortunately, soon after their discovery the grain boundaries in these new materials were observed to block current, similar to rare-earth barium cuprate (REBCO) materials like

5,14 YBa2Cu3O7-x (YBCO), but to a somewhat lesser extent. Remarkably, fine-grain, randomly oriented K-doped BaFe2As2 (Ba122) has been synthesized with global critical current density

-2 6,42 around 10 kAcm (4.2 K, 10 T) and textured tapes of K-doped Ba122 and SrFe2As2 (Sr122)

15,92 have now been produced that raise Jc by another order of magnitude. Despite these advances, applications are still dominated by the use of more mature superconducting materials. Here we demonstrate the potential use of bulk Ba122 magnets by synthesizing and characterizing the first iron-pnictide superconducting magnet capable of trapping over 1 T at 5 K. Electromagnetic and mechanical measurements suggest that the material properties are suitable for making larger bulk magnets that can be magnetized to trap very strong magnetic fields (> 10 T).

Conventional permanent magnets are limited by their magnetization saturation, and are therefore not capable of producing fields much greater than 1 T. However, induced persistent currents can be trapped inside a superconductor to produce magnetic fields (Btrapped) that scale with the size of the current loops flowing in the bulk:

= bulk where A0 is a geometrical factor, μ0 is the permeability of vacuum, Jc is the bulk or globally circulating critical current density, and r is the radius of the sample. High field bulk magnets bulk require high Jc and large r with a well-defined geometry. The field trapping ability is then 2 limited mostly by Jc(H) (Hc2 > 90 T for Ba122) and mechanical strength. Currently, mechanically reinforced REBCO materials produce record fields (> 17 T)125,126 at modest temperatures (> 20 K), but are limited in size (r ≤ 50 mm) because grain boundaries (GBs) block

current forcing such samples to be grown as single crystals. In contrast, MgB2 is not subject to intrinsic current blocking127 and can be manufactured as large diameter polycrystalline bulks,128–

131 but currently MgB2 lacks high enough Hc2 (limiting Jc(H)) to compete with YBCO for high

field applications. Despite its lower Tc and Jc than YBCO and MgB2, Ba122 has the geometric

versatility of MgB2 with better Jc(H) characteristics in high fields.

8.2 – Experimental Details Ba, K, Fe, and As were combined in a molar ratio of 0.6:0.42:2:2 and reacted together by a mechanochemical reaction followed by sintering in a hot isostatic press (HIP) at 600 °C. Details of the synthesis of the bulk material can be found elsewhere.42 After the two HIP heat treatments and subsequent re-milling, approximately 3-5 g of Ba122 powder was pressed into 15.9 mm

diameter pellets and then further densified in a cold isostatic press (CIP) at 276 MPa. These were then wrapped with Ag foil and inserted into a steel tube that was carefully machined to the diameter of the pellet + foil and both ends were plugged and welded shut under vacuum. The plugs were chamfered towards the pellet to help the steel tube compress around the pellet. The welded tubes were then swaged and CIPped as above, reducing the diameter of the samples ~10%. Finally, the samples were heat treated again for 10 hours at 600 °C in the HIP. After heat treatment, the steel tubes were sliced with a diamond saw to reveal the pellet surfaces. Several room-temperature Vickers hardness (HV) measurements were made on the surface of the pellets using loads from 25 g to 2000 g. Light and scanning electron microscopy was used to study and measure the micro-indentations.

Magneto optical (MO) imaging was used to image the local field profile produced by magnetization currents induced by field-cooling into the superconducting state in 120 mT applied perpendicular to the bulk sample’s surface and then removing the magnetic field. Due to the limited size of the cryostat, MO imaging was done on a 3.7 mm thick sample. Then, disk- shaped 122 bulks with ~10 mm in diameter and ~18.4 mm in total thickness were vertically stacked on either side of a spacer containing a transversal cryogenic Hall sensor to measure the magnetic flux density between the pellets (see inset to figure 8.2). Another Hall sensor was mounted to the outside end of the stack. The stack was cooled to ~5 K by a GM cryocooler under

an external field (Happ) of 8 T, and the external field was subsequently removed. After the field- cooling magnetization and reduction of the external field to zero, the magnetic flux density trapped in the bulk was measured at the center of the spacer as a function of increasing temperature (0.2 K/min) and separately as a function of time. For the magnetic hysteresis loop measurement, the sample was zero-field cooled to 5 K and its flux density in the sample stack was recorded as a function of increasing and decreasing external field.

8.3 – Results To assess the mechanical properties of the Ba122 bulks, room temperature Vickers Hardness tests were performed. The average HV was 3.5 (± 0.2) GPa. Cracks were observed propagating from the corners of the indentations, which is typical of brittle materials. Figure 1a shows an optical image of one of the bulks with thickness = 3.7 mm. The magnetic flux density produced by bulk current circulating over the sample is visualized in figure 1b and 1c by magneto optical

Figure 8.2 - Trapped field as a function of increasing temperature for the bulk sample stack that was field-cooled magnetized at 5 K. Simplified schematic of sample and Hall probe arrangement.

imaging. Figures 8.1(b) and 8.1(c) show remanent magnetic flux (Happ = 0) from induced currents that were trapped by applying magnetic fields of 120 mT and then field-cooling the

sample from above Tc to 11 K and 20 K, respectively, followed by removal of the applied magnetic field.

Figure 8.2 shows the trapped field (at Happ = 0) measured by Hall sensors placed on the surface

(H1) and between (H2) the stack of Ba122 bulk magnets after field cooling (Happ = 8 T) as a function of increasing temperature. It should be noted that H2 was placed 3.7 mm away from H1, which is 5.5 mm from the center of the stack where the maximum field would be expected. At 5 K, the bulk stack trapped 0.68 T on the outer surface (H1) and 1.02 T between the bulks

(H2). The trapped field decreased with increasing temperature and vanished at Tc ~33 K. The average macroscopic current density, which was estimated by a Biot-Savart approximation130 using the total thickness of the magnet stack and the experimentally obtained trapped field from -2 H1, is ~ 50 kAcm . This matches Jc obtained by local magnetization measurements made on 42 small bulk samples at T = 4.2 K and Happ = 0.6 T.

On field cooling, no magnetic flux jumps were observed at a ramp rate of 1.8 T/h. In one instance, an unexpected quench of the magnetizing magnet resulted in a rapid (~1 second)

removal of external field from Happ = 1.5 T to 0 T, corresponding to a ramp rate of > 2 T/sec. Despite the sudden removal of flux, the trapped field value quickly shifted to the expected critical state and magnet creep behavior was identical to data taken during the controlled process where Happ was slowly removed.

Figure 8.3 shows the magnetic flux density inside the stack of the initially zero-field cooled bulk samples as a function of increasing and decreasing external field. The samples showed good shielding behavior below Happ = 0.7 T and the hysteresis loop remained open at Happ > 8 T (see inset), which was the maximum applied field of our magnetizing magnet. Figure 8.4(a) shows the time dependence of trapped field at 5 K. The trapped field decayed approximately 3% after one day at 5 K. Figure 8.4(b) shows the normalized relaxation of trapped field as a function of time measured at 5 K, 10 K, and 20 K. The magnet creep rate increased with increased holding temperature due to thermally activated flux depinning. The trapped field decayed ~7 % after 1 day at 20 K.

Figure 8.4 - Trapped magnetic field magnet creep at H2. (a) Time dependence of trapped field at 5 K. (b) Normalized magnetic field creep as a function of time at 5, 10, and 20 K.

8.4 – Discussion The bulk magnets were processed by a scalable and versatile low-cost technique using milling, CIPping and HIPping, which are common ceramic processing techniques used in industry. The powder-in-tube technique and subsequent low-temperature reaction allow for several bulks to be produced in a single batch. They can then be sliced to a desired thickness, with the length and diameter of the HIP limiting the maximum bulk dimensions. In addition, the steel tubing adds a reinforcing ring that can easily be designed to further improve the mechanical strength of the bulk, as has proven invaluable for trapping high fields using REBCO bulks.126 High strength bulks and external reinforcement are important because the interaction between trapped field and current results in mechanical force that is proportional to Jc x B. Due to their brittle failure

mechanics, superconductors with high fracture toughness (KC) are desired for high field 0.5 applications. A KC of ~2.35 (± 0.14) MPa m was roughly calculated from the length of micro- cracks propagating from the corners of the micro-indentations according to the following formula:132

= 0.0726 / where P is loading force in N, C is the distance between center of indention and the tip of the crack in m, and 0.0726 is a calibration constant. This fracture toughness exceeds single crystal 133 134 135 Mn-doped Ba122, HIPped MgB2, bulk top-seeded melt-grown YBCO, and is about equal

136 to polycrystalline Al2O3. Table 8.1 summarizes the properties of Ba122 compared to YBCO

and MgB2 bulks.

The flux distribution observed by MO imaging indicates that the current trapped in the bulks is distributed macroscopically around the entire bulk sample. No flux avalanches were observed by Hall measurements during the magnetization process, even when the magnetizing magnet quenched, as discussed above, which suggests Ba122 is less susceptible to flux jumps and avalanches than MgB2 and YBCO bulks. This may be attributed to high thermal conductivity in the metallic Ba122 bulks or good thermal stability because of the small sample size tested here.

While high Hirr, supported by figure 3, suggests potential use for Ba122 bulks to trap high

magnetic fields, the magnet creep rate (~3% after a day at 5 K) is still higher than in MgB2 (~1.5% at 20 K) 129.

42,137,138 Figure 8.5 (a) shows a comparison of Jc(H) data taken from the literature. Figure 8.5 (b) trapped shows a calculation of maximum B as a function of r for K-doped Ba122 and MgB2 using the bulk Jc(H) data presented in figure 8.5 (a). This calculation is for an infinitely long cylinder and takes the radial field dependent Jc into account. In the infinite long cylinder geometry (A0 = 1), local current density j(x) varies with respect to radial direction x (0 < x < r) due to self-field but does not change with respect to circumferential and length directions. Thus, local flux density b(x) and j(x) can be calculated as,

() = () , () = ()

trapped The maximum B at the center of a bulk is given as b(r). We included Jc data for C-doped

MgB2 bulk material, though MgB2 bulk magnets in the literature are typically undoped and therefore have not demonstrated the ability to trap the high fields suggested by figure 8.5 (b) at

4.2 K. While MgB2 outperforms Ba122 at low fields and small r, Ba122 bulks are very competitive above the size limit of YBCO (r > 50 mm) at low temperatures, and samples with r ≥

Figure 8.5 - Comparison between K-doped Ba122 and MgB2. (a) Critical current density vs. 42 138 137 applied magnetic field for Ba122, undoped MgB2, and C-doped MgB2 bulks. Dotted lines are extrapolated data. (b) Maximum trapped field vs. radius for K-doped Ba122 and MgB2 polycrystalline bulks calculated from the data in (a) for an infinite thickness cylinder.

92 mm would be capable of trapping higher fields than MgB2, even at 20 K. Such large bulks would be useful for magnetic levitation in energy storage applications and could provide high fields in compact magnetic resonance devices.

The trapped field at the surface of the magnetized stack of Ba122 bulks as measured by H1 was 0.68 T. The magnitude of the field at the center of the stack should be about twice this value ~1.36 T. H2, which measured 1.02 T, was located between H1 and the center of the stack, and is in-between 0.68 and 1.36 T. The expected central trapped field ~1.36 T, given the radius ~5 mm, is about 65% of the ideal maximum trapped field presented in figure 8.5 (b), likely due to the finite thickness of the sample and slightly lower Jc(H) than shown in figure 8.5 (a).

While high-field applications for Ba122 using much larger bulk samples look very promising, our current Ba122 bulks may not be very competitive against MgB2 at low-fields due to their higher magnetic creep rate and lower Jc at low-fields. However, given improved Jc properties of Ba122 and Sr122 material reported in the literature87,92,139, there is still a lot of room to improve

Jc by better processing and partial alignment of grains. It is important to note that the values in figure 8.5 (b) are proportional to Jc(H), so modest improvements in Jc(H) are all that is needed for Ba122 to be very competitive with MgB2 even at low fields.

Table 8.1- Comparison of superconducting and mechanical properties for YBCO, MgB2 and K-doped Ba122

Material Form Tc Hc2(0 Hardness Fracture Limiting (K) K) (T) (GPa) toughness (MPa size (mm) m0.5) YBCO Single 92 >100 7-8135 1.4-1.6135 r ~ 50 crystal

140 134 134 MgB2 Polycrystal 39 ~ 30 10-12 1.3-1.4 no limit Ba122 Polycrystal 38 ~ 902 3.5 (± 0.2) 2.35 (± 0.14) no limit

8.5 – Conclusions In summary, we successfully synthesized the first bulk iron-pnictide demonstration magnet capable of trapping over 1 T (5 K) and 0.5 T (20K) by using fine-grain polycrystalline material and a scalable technique that could generate much larger samples. Magneto optical imaging suggests the material has macroscopic currents circulating throughout the entire sample. The time dependence of the trapped field showed a low magnetic creep rate (~3% after 24 hours at 5 K). Vickers hardness indentations indicate that the bulk material has a hardness ~3.5 GPa and a fracture toughness ~2.35 MPa m0.5. Larger bulks are expected to trap even higher fields, given

the high Hc2 of K-doped BaFe2As2. Modest improvements to Jc(H) will make Ba122 bulks very competitive against REBCO and MgB2 for bulk magnet applications.

CHAPTER NINE

EXPLORATION OF PATHS TOWARD HIGHER CRITICAL CURRENT DENSITIES IN K-DOPED BaFe2As2 POLYCRYSTALS

In the following chapter, Jianyi Jiang carried out some of the vibrating sample magnetometer measurements. Benjamin Hainsey helped synthesize some of the samples. Eric Hellstrom and David Larbalestier directed the research.

Beyond improving connectivity of grains and eliminating secondary phases that block current,

there are still a few ways to improve the pinning properties of the material and thus increase Jc and the irreversibility field (Hirr). The addition of defects and/or development of specific

micro/nano-structures to improve the maximum pinning force (Fp) is well documented for

superconductors and a common route to improve Jc and Hirr beyond the values obtained from single crystals. Here we tried to improve Fp in K-doped BaFe2As2 superconducting wires and bulks by three mechanisms: doping with non-superconducting impurities, doping with multiple global dopants, and over-doping. Each technique did not significantly improve Jc , through small

additions of Ag seemed to improve Jc at fields < 5 T. Jc was largely unaffected by overdoping of

K, despite decreasing Tc with K concentration. Based on these results, bulk transport is likely determined mostly by the GB properties at high fields that are not successfully improved by these techniques.

9.1 – Introduction Our Co-doped thin film results show that the material can support a large volume of non- superconducting impurity phases without significant degradation of superconducting properties, 36,37,41 and in doing so Jc can be improved by an order of magnitude. We have already made pulsed laser deposition targets with BaO2 added for thin film work at University of Wisconsin to artificially engineer such defects.141 For these impurities to be effective pinning sites, they must have dimensions on the order of twice the superconducting coherence length (2.5-7 nm) and be uniformly dispersed. One advantage of our synthesis technique is that these impurities can be mechanically alloyed into the material. This is a common technique that has been used to add dislocation pinning sites into various metal alloys on similar length scales to improve their material properties.142 Impurities that we attempted to incorporate coherently into the AE122

87 bulks include refractory/precious metals, and stable non-superconducting materials from the literature with similar crystal symmetry as AE122.

Variation in the superconducting order parameter on nm length scales can also provide vortex pinning sites. For example, a slight substitution of one AE cation for another can produce local regions with lesser superconducting properties, as can the slight substitution of one aliovalent dopant for another (ie. a partial replacement of K with Na, Co with Ni, or As with P). This kind of pinning mechanism may be better than introducing impurities because the pinning sites are still superconducting and do not completely block current when the vortex density is less than the pin density. Since vortex density changes with field and temperature, Fp may be improved over a wider range of conditions whereas impurity pins may only be effective for specific temperature and matching field ranges. In addition to this electronic doping mechanism, there may be an additional structural source of pinning caused by a difference in atomic radius of the dopant atoms that may be on a different length scale.

Lastly, if grain size is sufficiently small, GBs can also serve to pin vortices. This is the primary 143 global pinning mechanism for Nb3Sn. Jc in this case has been shown to be inversely proportional to the area of GB per unit volume.144 This is because although superconducting properties at the grain boundaries may be depressed, grain boundaries are not heavily intrinsically weak linked and so a higher density of grain boundaries is advantageous to improve Fp.

9.2 – Experimental Details

K-doped BaFe2As2 samples were synthesized via a mechanochemical reaction path and subsequent 600 °C reaction as reported elsewhere.42 10 samples were synthesized by combining the elements according to their stoichiometry as follows: (Ba0.6K0.4)Fe2As2, (Ba0.6K0.42)Fe2As2,

(Ba0.6K0.44)Fe2As2, (Ba0.5K0.5)Fe2As2, (Ba0.4K0.6)Fe2As2, (Ba0.6K0.4)Fe2(As0.95P0.05)2,

(Ba0.6K0.2Na0.2)Fe2As2, (Ba0.6K0.35Na0.05)Fe2As2 , (Ba0.6K0.42)Fe2As2 + 4wt% Sn, and

(Ba0.6K0.42)Fe2As2 + 4wt% Ag. Five samples had impurities added after the first heat treatment, but before the second milling and powder consolidation steps to obtain samples with

(Ba0.6K0.42)Fe2As2 + 4wt% Sn, (Ba0.6K0.42)Fe2As2 + 10wt% Sn, (Ba0.6K0.42)Fe2As2 + 4wt% Ag

(Ba0.6K0.42)Fe2As2 + 10wt% Ag, and (Ba0.6K0.42)Fe2As2 + 10wt% BaZrO3 (BZO). A magnetic property measurement system (Quantum design: MPMS-XL5s) was used to measure the

magnetization as a function of increasing temperature after zero-field cooling (ZFC) the sample to 5 K and applying a magnetic field of 20 Oe. Magnetization measurements were made using a 14 T Oxford vibrating sample magnetometer (VSM) with the magnetic field parallel to the

sample’s length. Jc was calculated from the VSM magnetic hysteresis measurements as described in section 2.4.2.

9.3 – Results

In this study, we use two main figures of merit to compare samples. Firstly, we assess the Tc of a material by measurements of volumetric susceptibility (χ) as a function of increasing temperature. The point at which the sample is no longer diamagnetic defines Tc, and the sharpness of the transition tells us if there is a broad range of Tc variability within the bulk sample. In addition, the shielding fraction (SF) (SF -χ) tells us whether is proportional to magnetic flux is completely shielded from the material or if it can penetrate some of the sample due to poorly- or non-superconducting regions. However, the magnitude of χ contains a large amount of experimental error (± 8%) for bulks as discussed in section 2.4.2. Secondly, we assess

the critical current density (Jc) as a function of applied magnetic field (μ0H) calculated from magnetization measurements assuming the magnetic response comes from the entire sample. As discussed in chapter 6, this is a valid assumption for these fine grained samples since the contribution of the intragranular magnetization to the measurement is small. Figure 9.1(a) and

9.1(b) show χ(T) and Jc(μ0H) respectively for samples with either 0, 5, or 10 at. % extra K. Little dependency on properties with respect to added K is observed.

Figure 9.2 shows how χ(T) and Jc(μ0H) vary for samples that were over-doped containing either

40, 50, or 60 at. % K substituted for Ba. Figure 9.2(a) shows Tc decreases systematically from 35.9 K in the optimally (40% K) doped sample to 29.3 K in the sample with 60 % K. In

addition, figure 9.2(b) shows the optimally doped sample has the highest Jc, though Jc(H) behavior does not vary by much.

Figure 9.3 shows χ(T) and Jc(μ0H) for various samples with multiple dopants. As seen in figure

9.3(a), the addition of multiple dopants lowers Tc. In addition, the diamagnetic response of the sample containing both K at the Ba site, and P at the As site was suppressed. Figure 9.2(b) shows Jc(μ0H) behavior for the samples with multiple dopants. All samples had lower Jc(μ0H) at

most fields compared to the standard 40% K-doped control sample. The sample with slight (~14

at. %) Na substituted for K, had a broad Jc(μ0H) dependence at low fields.

Figure 9.4 shows χ(T) and Jc(μ0H) for samples in which small amounts of impurity phases were introduced. Ag, Sn, and BZO were added before final milling and heat treatment, except samples labled “in situ”, in which the impurities were added with the elements at the very beginning of the synthesis procedure. All samples in figure 9.4(a) showed degraded Tc

compared to the control sample, and samples containing Sn had less of a diamagnetic response. In situ samples showed nearly identical χ(T) behavior to samples with the same composition in

which the impurities were added later during processing. Figure 9.4(b) shows Jc(μ0H) for some of the samples with impurities added. Samples with 10 wt% Sn and 10wt% BZO had lower Jc than the control sample while addition of 4 wt% Ag improved the low field Jc(μ0H) behavior.

Table 9.1 – Superconducting properties of samples measured

-2 -2 T (K) J (kAcm ) J (kAcm ) Composition c c c (χ = -0.1) (4.2 K, 0 T) (4.2 K, 10 T)

(Ba0.6K0.4)Fe2As2 35.8 114.8 9.5

(Ba0.6K0.42)Fe2As2 35.0 128.2 8.6

(Ba0.6K0.44)Fe2As2 35.5 109.9 7.8

(Ba0.5K0.5)Fe2As2 33.8 99.0 7.5

(Ba0.4K0.6)Fe2As2 29.3 94.5 6.9

(Ba0.6K0.35Na0.05)Fe2As2 32.3 50.7 4.3

(Ba0.6K0.4)Fe2(As0.95P0.05)2 33.5 123.5 7.5

(Ba0.6K0.2Na0.2)Fe2As2 34.0 61.8 4.3

(Ba0.6K0.42)Fe2As2 + 4 wt% Sn in situ 27.2

(Ba0.6K0.42)Fe2As2 + 4 wt% Sn 27.6

(Ba0.6K0.42)Fe2As2 + 10 wt% Sn 25.9 27.1 1.3

(Ba0.6K0.42)Fe2As2 + 4 wt% Ag in situ 33.7 132.3 8.5

(Ba0.6K0.42)Fe2As2 + 4 wt% Ag 33.8

(Ba0.6K0.42)Fe2As2 + 10 wt% Ag 30.9 72.7 4.2

(Ba0.6K0.42)Fe2As2 + 10 wt% BZO 33.7 54.6 4.3

9.4 – Discussion 65 64,145–149 Irradiation of Sm1111 and doped BaFe2As2 single crystals has enabled the creation of local -2 nm sized point defects improving Jc several times up to 10 MAcm for K-doped BaFe2As2 (SF, 4.2K).146 Similar attempts at adding defects in our bulk material by irradiation have not been global 150 as successful in improving Jc . However, effective pinning sites can also be engineered into a material during growth by including non-superconducting impurity phases or by inducing defects. Addition of extra K is a common practice amongst groups that synthesize K-doped 15,61,83,84,87,92,151 BaFe2As2 or SrFe2As2 superconductors. The reason for adding extra K is

typically to compensate for loss of K during synthesis, but Wang et al. reported on enhanced 84 Jc(μ0H) properties by overdoping that they attributed to improved intrinsic pinning. In our case, addition of extra K appears to be ineffective at improving Jc, likely because our closed process does not easily lose K and our lower temperature reaction already results in a high defect density for our samples that is not further improved by additional K. Based on our direct observation of K deficiency across GBs, reported in chapter 7, we hypothesized that the GB composition could be pushed closer to optimum K-concentration by overdoping the entire sample. The results in figure 9.2 show that while we were successful in overdoping our material,

Jc was not enhanced.

By substituting multiple dopants, we hoped to improve pinning properties by introducing random defects that could act as pins for vortices, or pins for GBs during grain growth to decrease the

overall grain size. As seen in figure 9.3, no improvement of Jc was obtained. Similarly, we tried the addition of impurity phases that had been reported to be successful by other groups in raising global Jc in their polycrystalline materials. Many have reported the addition of Sn was successful in increasing grain connectivity,15,73,152 though Sn additions showed the greatest detrimental effects for our samples suggesting that Sn addition is not applicable for our low-temperature synthesis route. However, the addition of a small amount (4 wt. %) of Ag was successful in improving global Jc at low fields compared to the control sample, though the effect does not extend to the high field regime. For two samples, Sn or Ag was added during different parts of the reaction path. The impurities were introduced both before (in situ) and after the first heat treatment, with

almost no change in superconducting properties. Presently, the highest obtained Jc values for pnictide tapes contain Ag or Sn addition.15,87 We also attempted to incorporate BZO nanopowder (grain size ~ 40 nm) into samples to attempt to mimic Jc enhancement observed in thin films with BZO addition,153,154 but this was unsuccessful.

Overall, we reason that two scenarios likely occur that explain our results. Either the intrinsic pinning within our crystallites is already optimized via our low temperature reaction path that results in a high density of defects (as has been observed by transmission electron microscopy),6 or with any degradation of crystal quality resulting in improvement in the intrinsic Jc(μ0H ) properties of these superconductors, there is a concurrent degradation of GB properties effectively increasing the Josephson contact thickness. In this second scenario, the overall

Josephson current density across GBs is proportional to the intrinsic Jc, but inversely dependent on the thickness of the contact.56,155

During these experiments we became aware of issues concerning sample stability between measurements. While χ(T) varies little over time, Jc(μ0H) can decrease by a factor of two (for optimally-doped) and three (for over-doped) over 11 weeks. Most samples were measured within a couple weeks of synthesis, but we can’t for certain quantify the effect of degradation on our measurements between samples that were measured at different times. While we were not aware of the stability of samples at the time of these experiments, the primary goal has been to

increase Jc by at least 5-10 times, which we are confident, has not been accomplished by these methods.

9.5 – Conclusions In summary, we attempted to raise the intergranular critical current density my introducing

dispersed defects in bulk polycrystalline samples. Attempts to raise intergranular Jc by adding extra K, overdoping, double doping, and introducing impurity phases were unsuccessful, except for the addition of slight amounts of Ag that improved low field Jc performance. We reason that addition of random defects either does not improve the intrinsic pinning of our already highly disordered material, or concurrently results in additional degradation of the weak-linked grain boundaries that limit bulk current transfer.

APPENDIX A

PROCESSING IRON BASED SUPERCONDUCTORS SAFELY

Eric Hellstrom, Jeremy Weiss Updated: Feb. 19 2013

The literature describing the synthesis of LaFeAsO, BaFe2As2, and other iron based superconductors uses solid-state synthesis with a variety of starting materials. The starting materials are ground together, pressed, placed in a quartz tube, sealed, and reacted at temperatures up to 1250°C.

The potential hazards are the toxicity of As (arsenic), and the possibility of an explosion when reacting the chemicals in a sealed quartz tube. As is best known as rat poison and as the poison of choice in the play Arsenic and Old Lace. But it is present in materials that have important properties in the modern world, including all red and infrared light emitting diodes (LED) that are taillights in high-end cars and the LEDs in stop lights, lasers in laser pointers, CDs and DVD players, and in power amplifiers in cell phones. The general public does not worry about, and indeed in most instances does not even know about, the toxicity of As in these compounds.

We plan the following safety procedures. Integrated Safety Management is actively implemented and maintained by each individual working in the laboratory. Please see the “Five Core Functions for the Synthesis of Iron Based Superconductors.”

We will continue to synthesize BaFe2As2, doped BaFe2As2, and other related As compounds in which Ba is replaced with other alkali and alkaline earth elements using solid state synthesis. The procedure we will follow is illustrated by the reaction to form the parent compound

BaFe2As2.

Ba + 2Fe + 2As = BaFe2As2.

The Ba, Fe, and As will be weighed out using a standard balance. This will be done in a glove box with foil covering the area around the balance to catch any As powder that may be spilled.

The As, Ba, and Fe will be ground using a mortar and pestle or a SPEX mill that is located inside of a glovebox. Cleaning of equipment possibly contaminated with As will be wiped clean in the glove box and then transferred it in a plastic bag from the glove box into a fume hood. Acid resistant gloves and goggles will be worn when cleaning. A dilute (approx. 20-30%) nitric acid

in H2O solution will be used to dissolve contaminated ceramic material, and disposed of in an acid waste container stored under the fume hood. All wipes used to aid in cleaning will be disposed of as hazardous waste.

Any time someone handles material containing As, they will wear latex gloves. These will be disposed of after each use in a plastic container labeled indicating it contains hazardous waste. Laboratory items, such as spatulas, that are in contact with As will be wiped with Kim Wipes to remove excess powder, and then washed with soap and water. Powders will only be handled in fume hoods or the glove box. The Kim Wipes and any excess powder will be disposed of as hazardous waste.

After grinding the elements together, the powder will be pressed into pellets using a plunger and die set inside the glove box or wrapped in foil, packed into a metal tube which will be evacuated of air and welded shut.

We do not anticipate any explosions, but want to be prepared in case one does occur.

Heat treatments are carried out at temperatures up to 1250 °C in stainless steel or niobium tubes that have been evacuated of air and welded shut. A hot isostatic press (HIP) provides a positive pressure of 28 ksi to suppress the sublimation of elemental As during the heat treatment. The HIP contains many safety features including a closed, isolated, ASME approved pressure chamber that vents outside. In the unlikely event that a sample leaks in the HIP chamber, any As vapor will condense on the chamber walls. The chamber will then have to be decontaminated and cleaned by someone wearing an OSHA approved respirator and gloves.

We have successfully used this technique to synthesize hundreds of As-containing samples without leakage during the heat treatment.

Fully reacted samples will be stored in glass or plastic sample bottles, marked with the chemical composition. These will be stored in the laboratory.

We anticipate that solid pieces of BaFe2As2 rather than powders will be used for almost all the electromagnetic measurements at NHMFL. These are easier to handle than powder. Anyone who measures samples will be instructed to wear latex gloves when handling the material and dispose of soiled gloves as hazardous waste.

Initial removal of samples from the stainless steel ampule will be done in the fume hood with tin snips or in the high speed diamond saw under oil that is disposed of as hazardous waste. Pieces of the fully reacted sample will be cut for the electromagnetic studies. This will be done with the low speed diamond saw in the fume hood. Powder residue from cutting samples will be disposed of as hazardous waste. Any grinding to shape the sample will also be done in the fume hood using SiC paper at low speeds to contain powder residue. The SiC will be disposed of as hazardous waste.

Wires will be made by the standard powder in tube method. Metal tubes are filled with powder inside the glovebox. The tubes are plugged and the plugged end is swaged shut to create a mechanical seal. The sealed tubes are then successively swaged, rolled, and drawn into wire. The wire is then sectioned in the glovebox and welded into crucibles for heat treating.

APPENDIX B

INTEGRATED SAFETY MANAGEMENT PLAN

Scope of work

• BaFe2As2, doped BaFe2As2 and other As-containing compounds are synthesized from the elements and characterized.

Hazards • As is poisonous, especially when oxidized. • As sublimes at 614 °C creating an explosion hazard when heated.

• As can form AsH3 gas in the presence of a reaction the produces monatomic H atoms.

AsH3 is colorless and odorless with lethal doses below 100 ppm • Ba and Fe metals are flammable.

Hazard control • All handling of elements is done in a glovebox specifically designated for use with hazardous materials which is kept in a locked room for hazardous materials research only. • All MSDS for hazardous materials are kept next to the door in said room. • All people that use the hazardous materials room are trained by their supervisor as to the specific risks of working with hazardous materials, the location of personal protective equipment (PPE), and the proper use of PPE. • A respirator for hazardous chemical dust is kept in a visible, designated place in the room. • All people working in the hazardous materials room have read the World Health Organization’s IPCS health and Safety Guide No. 70 on inorganic arsenic compounds. A copy is located in the room with the MSDSs. • Transferring of hazardous materials is done in air-tight containers and at least two containers are used in case one fails.

• Heat treatments are carried out at temperatures up to 1250C in stainless steel or niobium tubes that have been evacuated of air and welded shut. A hot isostatic press (HIP) provides a positive pressure of 28 ksi to suppress the sublimation of elemental As during the heat treatment. The HIP contains many safety features including a closed, isolated, ASME approved pressure chamber that vents outside. • After reaction, the material is removed from the stainless steel using a diamond saw or tin-snips and handled in a fume hood. • Cutting, polishing, grinding and general handling of As-containing compounds is done in the fume hood with appropriate PPE or in a glove box if they are air-sensitive. Polishing waste is disposed of as hazardous waste. All polishing is dry polishing to prevent possible As runoff waste. • As-containing samples are kept in air-tight containers and kept in a dry-box for long-term storage. Air-sensitive material is stored in a bottle in a glovebox or canning jar under Ar gas (double contained). • Cleaning of Equipment and tools that are possibly contaminated by As is done in fume hoods wearing latex or nitrile gloves and all wipes, paper towels, foil, gloves, and excess material are disposed of as hazardous waste • Dilute Nitric Acid solution (20-30% acid) used for cleaning mortars, pestles, and milling jars is handled with appropriate PPE (Goggles and acid resistant gloves) and disposed of as liquid hazardous waste in the fume hood. • GOOD HOUSEKEEPING is maintained!

Perform work within controls • So far controls have successfully prevented any injuries as a result of accidents. • Employees are asked to constantly consider ways to further improve procedures.

Feedback and improvement • Air-tight, heavy walled stainless steel containers to replace glass-canning jars and plastic bags for transferring hazardous material (currently being machined). • Installation of new fume hoods in C323 to keep Arsenic work localized.

APPENDIX C

PRODUCTION OF MULTIFILAMENTRY K-DOPED WIRES

Multifilament wires were processed out of (Ba0.6K0.4)Fe2As2 polycrystalline material. 7, 37, and 259 filamentary wires were produced by drawing and restacking processes. The 7-filament wire with filament diameter d ~ 120 μm had transport Jc comparable to mono-filamentary wires. The

37 filament wire with d ~ 95 μm had degraded Jc by over a factor of two compared to mono- filamentary wires while the 259 filament wire with d ~ 25 μm had no Ic and severely degraded

Tc. Based off scanning electron microscopy and energy dispersive x-ray spectroscopy (EDS), we determine that our Ag diffusion barrier, the thickness of which decreased with decreasing filament size, was ineffective at protecting the bulk superconducting material from reacting with the Cu sheath.

Figure C.1 – Images of transverse cross section of various wires with different filament counts and sizes.

100

Figure C.2 – Volumetric susceptibility of K-doped BaFe2As2 wires with various filament counts as a function of increasing temperature after cooling in zero field and then applying 2 mT.

Figure C.3 – (left) Transport critical current density as a function of applied field and (right) corresponding n-values calculated from the I-V curves.

Figure C1 shows images of the transverse cross section of each wire studied. EDS analysis showed the fine filaments formed secondary phases containing large quantities of cu. Figure C2 shows the zero field cooled susceptibility as a function of increasing temperature for each wire showing that Tc is decreasing with increasing filament count and decreasing filament diameter

101

transport and Ag layer thickness. Figure C3 shows Jc as a function of applied field and corresponding n-values calculated from the I-V curves for the monocore, 7 and 37 filament wires. The degradation of n-value is likely also due to Cu contamination of the superconductor.

In summary, multifilament wires were processed using K-doped BaFe2As2 superconducting material inside a Ag and Cu sheath and electromagnetic characterization was performed. A decrease of Tc and broadening of the Tc susceptibility occurred with finer filament size. Jc and n- values showed a similar trend. Cu was found to contaminate wires with very fine filaments, resulting in the degraded behavior.

102

APPENDIX D

Dear Jeremy Weiss, Thank you for your request to reproduce IOP Publishing material.

Figure 10 Supercond. Sci. Technol. 27 (2014) 044002 to be reused in your dissertation to be published by UMI Company/ ProQuest.

We are happy to grant permission for the use you request on the terms set out below. Conditions Non-exclusive, non-transferrable, revocable, worldwide, permission to use the material in print and electronic form will be granted subject to the following conditions: • Permission will be cancelled without notice if you fail to fulfil any of the conditions of this letter. • You will make reasonable efforts to contact the author(s) to seek consent for your intended use. Contacting one author acting expressly as authorised agent for their co- authors is acceptable. • You will reproduce the following prominently alongside the material: o the source of the material, including author, article title, title of journal, volume number, issue number (if relevant), page range (or first page if this is the only information available) and date of first publication. This information can be contained in a footnote or reference note; or o a link back to the article (via DOI); and o if practical and IN ALL CASES for works published under any of the Creative Commons licences the words “© IOP Publishing. Reproduced by permission of IOP Publishing. All rights reserved” • The material will not, without the express permission of the author(s), be used in any way which, in the opinion of IOP Publishing, could distort or alter the author(s)’ original intention(s) and meaning, be prejudicial to the honour or reputation of the author(s) and/or imply endorsement by the author(s) and/or IOP Publishing.

103

• Payment of £0 is received in full by IOP Publishing prior to use. Kind regards, Zora Catterick Publishing Assistant IOP Publishing Please note: We do not usually provide signed permission forms as a separate attachment. Please print this email and provide it to your publisher as proof of permission.

104

REFERENCES

1. Kamihara, Y., Watanabe, T., Hirano, M. & Hosono, H. Iron-Based Layered Superconductor La(O1-xFx)FeAs (x= 0.05−0.12) with Tc = 26 K. J. Am. Chem. Soc. 130, 3296–3297 (2008).

2. Tarantini, C. et al. Significant enhancement of upper critical fields by doping and strain in iron-based superconductors. Phys. Rev. B 84, (2011).

3. Fang, L. et al. High, magnetic field independent critical currents in (Ba,K)Fe2As2 crystals. Appl. Phys. Lett. 101, 012601 (2012).

4. Putti, M. et al. New Fe-based superconductors: properties relevant for applications. Supercond. Sci. Technol. 23, 034003 (2010).

5. Lee, S. et al. Weak-link behavior of grain boundaries in superconducting Ba(Fe1-xCox)2As2 bicrystals. Appl. Phys. Lett. 95, 212505 (2009).

6. Weiss, J. D. et al. High intergrain critical current density in fine-grain (Ba0.6K0.4)Fe2As2 wires and bulks. Nat. Mater. 11, 682–685 (2012).

7. Chen, G. F. et al. Transport and anisotropy in single-crystalline SrFe2As2 and A0.6K0.4Fe2As2 ( A=Sr , Ba) superconductors. Phys. Rev. B 78, (2008).

8. Chen, G. F. et al. Superconductivity at 41~K and Its Competition with Spin-Density-Wave Instability in Layered CeO1-xFxFeAs. Phys. Rev. Lett. 100, 247002 (2008).

9. Zhi-An, R. et al. Superconductivity at 55 K in Iron-Based F-Doped Layered Quaternary Compound Sm(O1-xFx)FeAs. Chin. Phys. Lett. 25, 2215–2216 (2008).

10. Yamamoto, A. et al. Evidence for two distinct scales of current flow in polycrystalline Sm and Nd iron oxypnictides. Supercond. Sci. Technol. 21, 095008 (2008).

11. Yamamoto, A. et al. Evidence for electromagnetic granularity in polycrystalline Sm1111 iron-pnictides with enhanced phase purity. Supercond. Sci. Technol. 24, 045010 (2011).

105

12. Durrell, J. H. et al. The behavior of grain boundaries in the Fe-based superconductors. Rep. Prog. Phys. 74, 124511 (2011).

13. Kametani, F. et al. Combined microstructural and magneto-optical study of current flow in polycrystalline forms of Nd and Sm Fe-oxypnictides. Supercond. Sci. Technol. 22, 015010 (2009).

14. Katase, T. et al. Advantageous grain boundaries in iron pnictide superconductors. Nat. Commun. 2, 409 (2011).

15. Zhang, X. et al. Realization of practical level current densities in Sr0.6K0.4Fe2As2 tape conductors for high-field applications. Appl. Phys. Lett. 104, 202601 (2014).

16. Sakagami, A. et al. Critical current density and grain boundary property of BaFe2(As,P)2 thin films. Phys. C Supercond. (2013). doi:10.1016/j.physc.2013.04.047

17. Hilgenkamp, H. & Mannhart, J. Grain boundaries in high-Tc superconductors. Rev. Mod. Phys. 74, 485–549 (2002).

18. Shimoyama, J. Potentials of iron-based superconductors for practical future materials. Supercond. Sci. Technol. 27, 044002 (2014).

19. Iida, K. et al. Epitaxial Growth of Superconducting Ba(Fe1-xCox)2As2 Thin Films on Technical Ion Beam Assisted Deposition MgO Substrates. Appl. Phys. Express 4, 013103 (2010).

20. Katase, T. et al. Biaxially textured cobalt-doped BaFe2As2 films with high critical current density over 1 MA/cm2 on MgO-buffered metal-tape flexible substrates. Appl. Phys. Lett. 98, 242510 (2011).

21. Manousakis, E., Ren, J., Meng, S. & Kaxiras, E. Is the nature of magnetic order in copper- oxides and in iron-pnictides different? Solid State Commun. 150, 62–65 (2010).

22. Scanlan, R. M., Malozemoff, A. P. & Larbalestier, D. C. Superconducting materials for large scale applications. Proc. IEEE 92, 1639–1654 (2004).

23. Kametani, F. et al. Intergrain current flow in a randomly oriented polycrystalline SmFeAsO0.84 oxypnictide. Appl. Phys. Lett. 95, 142502 (2009).

106

24. Wang, C. et al. Low-temperature synthesis of SmO0.8F0.2 FeAs superconductor with Tc= 56.1 K. Supercond. Sci. Technol. 23, 055002 (2010).

25. Graser, S. et al. How grain boundaries limit supercurrents in high-temperature superconductors. Nat. Phys. 6, 609–614 (2010).

26. Bennemann, K.-H. & Ketterson, J. B. The Physics of Superconductors: Vol. II: Superconductivity in Nanostructures, High-Tc and Novel Superconductors, Organic Superconductors. (Springer Science & Business Media, 2004).

27. Goodman, B. B. Type II superconductors. Rep. Prog. Phys. 29, 445 (1966).

28. Abrikosov, A. A. On the Magnetic properties of superconductors of the second group. Sov.Phys.JETP 5, 1174–1182 (1957).

29. Abrikosov, A. & Gorkov, L. Contribution to the Theory of Superconducting Alloys with Paramagnetic Impurities. Sov. Phys. Jetp-Ussr 12, 1243–1253 (1961).

30. Brandt, E. H. The flux-line lattice in superconductors. Rep. Prog. Phys. 58, 1465 (1995).

31. Blatter, G., Feigel’man, M. V., Geshkenbein, V. B., Larkin, A. I. & Vinokur, V. M. Vortices in high-temperature superconductors. Rev. Mod. Phys. 66, 1125–1388 (1994).

32. Ginzburg, V. L. On Superconductivity and Superfluidity (What I Have and Have Not Managed to Do), as well as on the‘Physical Minimum’ at the Beginning of the 21st Century. ChemPhysChem 5, 930–945 (2004).

33. Bean, C. P. Magnetization of High-Field Superconductors. Rev. Mod. Phys. 36, 31–39 (1964).

34. Prigozhin, L. The Bean Model in Superconductivity: Variational Formulation and Numerical Solution. J. Comput. Phys. 129, 190–200 (1996).

35. Itzler, M. & Tinkham, M. Vortex pinning by disordered columnar defects in large Josephson junctions. Phys. Rev. B 53, R11949–R11952 (1996).

36. Tarantini, C. et al. Strong vortex pinning in Co-doped BaFe2As2 single crystal thin films. Appl. Phys. Lett. 96, 142510 (2010). 107

37. Zhang, Y. et al. Self-assembled oxide nanopillars in epitaxial BaFe2As2 thin films for vortex pinning. Appl. Phys. Lett. 98, 042509 (2011).

38. Chen, X. H. et al. Superconductivity at 43 K in SmFeAsO1-xFx. Nature 453, 761–762 (2008).

39. Johrendt, D. & Pöttgen, R. Superconductivity, magnetism and crystal chemistry of Ba1−xKxFe2As2. Phys. C Supercond. 469, 332–339 (2009).

40. Rotter, M., Tegel, M. & Johrendt, D. Superconductivity at 38 K in the Iron Arsenide Ba1- xKxFe2As2. Phys. Rev. Lett. 101, 107006 (2008).

41. Lee, S. et al. Template engineering of Co-doped BaFe2As2 single-crystal thin films. Nat. Mater. 9, 397–402 (2010).

42. Weiss, J. D., Jiang, J., Polyanskii, A. A. & Hellstrom, E. E. Mechanochemical synthesis of pnictide compounds and superconducting Ba0.6K0.4Fe 2As2 bulks with high critical current density. Supercond. Sci. Technol. 26, 074003 (2013).

43. Ma, Y. et al. Fabrication and characterization of iron pnictide wires and bulk materials through the powder-in-tube method. Phys. C Supercond. 469, 651–656 (2009).

44. Wang, L. et al. Large transport critical currents of powder-in-tube Sr0.6K0.4Fe2As2/Ag superconducting wires and tapes. Phys. C Supercond. 470, 183–186 (2010).

45. Wang, L. et al. Low-temperature synthesis of SmFeAsO0.7F0.3−δ wires with a high transport critical current density. Supercond. Sci. Technol. 23, 075005 (2010).

46. Ma, Y. Progress in wire fabrication of iron-based superconductors. Supercond. Sci. Technol. 25, 113001 (2012).

47. Togano, K., Matsumoto, A. & Kumakura, H. Fabrication and transport properties of ex situ powder-in-tube (PIT) processed (Ba,K)Fe2As2 superconducting wires. Solid State Commun. 152, 740–746 (2012).

48. Pyon, S. et al. Enhancement of critical current densities by high-pressure sintering in (Sr,K)Fe2As2 PIT wires. Supercond. Sci. Technol. 27, 095002 (2014).

108

49. Togano, K. et al. Enhanced high-field transport critical current densities observed for ex situ PIT processed Ag/(Ba, K)Fe 2 As 2 thin tapes. Supercond. Sci. Technol. 26, 065003 (2013).

50. Togano, K., Gao, Z., Matsumoto, A. & Kumakura, H. Enhancement in transport critical current density of ex situ PIT Ag/(Ba, K)Fe2As2 tapes achieved by applying a combined process of flat rolling and uniaxial pressing. Supercond. Sci. Technol. 26, 115007 (2013).

51. Hellstrom, E. Viewpoint on fast track communication by Z Gao et al ‘high transport Jc in magnetic fields up to 28 T of stainless steel/Ag double sheathed Ba122 tapes fabricated by scalable rolling process’. Supercond. Sci. Technol. 28, 040501 (2015).

52. Gao, Z. et al. Superconducting properties of granular SmFeAsO1-xFx wires with Tc= 52 K prepared by the powder-in-tube method. Supercond. Sci. Technol. 21, 112001 (2008).

53. Zhang, X. et al. Effect of sheath materials on the microstructure and superconducting properties of SmO0.7F0.3FeAs wires. Phys. C Supercond. 470, 104–108 (2010).

54. Müller, K.-H., Andrikidis, C., Liu, H. & Dou, S. Intergranular and intragranular critical currents in silver-sheathed Pb-Bi-Sr-Ca-Cu-O tapes. Phys. Rev. B 50, 10218–10224 (1994).

55. D’yachenko, A. I. Hysteresis of transport critical current of high Tc superconductors in strong magnetic fields. Theory. Phys. C Supercond. 213, 167–178 (1993).

56. Svistunov, V. M., Aomine, T., Fukami, T., Yamamoto, T. & D’Yachenko, A. I. Hysteretic Phenomena in the Critical Current of High-Tc Superconductors. Mod. Phys. Lett. B 11, 1133–1151 (1997).

57. Rao, C. N. R. Concise encyclopedia of magnetic and superconducting materials, 1992, (ed.) Jan Evetts. Bull. Mater. Sci. 16, 85–85 (1993).

58. Welp, U., Xiao, Z. L., Novosad, V. & Vlasko-Vlasov, V. K. Commensurability and strong vortex pinning in nanopatterned Nb films. Phys. Rev. B 71, 014505 (2005).

59. Katase, T. et al. Atomically-flat, chemically-stable, superconducting epitaxial thin film of iron-based superconductor, cobalt-doped BaFe2As2. Solid State Commun. 149, 2121–2124 (2009).

109

60. Hiramatsu, H. et al. Microstructure and transport properties of [001]-tilt bicrystal grain boundaries in iron pnictide superconductor, cobalt-doped BaFe2As2. Mater. Sci. Eng. B 177, 515–519 (2012).

61. Ma, Y. Progress in wire fabrication of iron-based superconductors. Supercond. Sci. Technol. 25, 113001 (2012).

62. Matsumoto, A., Togano, K. & Kumakura, H. Microstructure and superconducting properties of Ag-sheathed (Ba, K)Fe2As2+Ag superconducting wires fabricated by an ex situ powder-in-tube process. Supercond. Sci. Technol. 25, 125010 (2012).

63. Wang, L. et al. Direct observation of nanometer-scale amorphous layers and oxide crystallites at grain boundaries in polycrystalline Sr1−xKxFe2As2 superconductors. Appl. Phys. Lett. 98, 222504 (2011).

64. Eisterer, M. et al. Disorder effects and current percolation in FeAs-based superconductors. Supercond. Sci. Technol. 23, 054006 (2010).

65. Eisterer, M. et al. Neutron irradiation of SmFeAsO1-xFx. Supercond. Sci. Technol. 22, 065015 (2009).

66. Yamamoto, A. et al. Small anisotropy, weak thermal fluctuations, and high field superconductivity in Co-doped iron pnictide Ba(Fe1−xCox)2As2. Appl. Phys. Lett. 94, 062511 (2009).

67. Yamamoto, A. et al. Evidence for electromagnetic granularity in the polycrystalline iron- based superconductor LaO 0.89F0.11FeAs. Appl. Phys. Lett. 92, 252501 (2008).

68. Leithe-Jasper, A., Schnelle, W., Geibel, C. & Rosner, H. Superconducting State in SrFe2- xCoxAs2 by Internal Doping of the Iron Arsenide Layers. Phys. Rev. Lett. 101, (2008).

69. Kim, Y. & Cho, Y. Synthesis of transition metal pnictide nanocrystalline powders by mechanochemical reaction. J. Alloys Compd. 393, 211–218 (2005).

70. Kim, Y. & Cho, Y. Magnetic transition of (MnFe)PAs prepared by mechanochemical reaction and post-annealing. J. Alloys Compd. 394, 19–23 (2005).

110

71. Takacs, L. Self-sustaining reactions induced by ball milling. Prog. Mater. Sci. 47, 355–414 (2002).

72. Takacs, L. Ball Milling-Induced Combustion in Powder Mixtures Containing Titanium, Zirconium, or Hafnium. J. Solid State Chem. 125, 75–84 (1996).

73. Gao, Z. et al. High transport critical current densities in textured Fe-sheathed Sr1- xKxFe2As2+Sn superconducting tapes. Appl. Phys. Lett. 99, 242506 (2011).

74. Larbalestier, D. C., Gurevich, A., Feldmann, D. M. & Polyanskii, A. A. High-Tc superconducting materials for electric power applications. Nature 414, 368–377 (2001).

75. Qi, Y. et al. Transport critical currents in the iron pnictide superconducting wires prepared by the ex situ PIT method. Supercond. Sci. Technol. 23, 055009 (2010).

76. Fujioka, M. et al. Effective ex-situ fabrication of F-doped SmFeAsO wire for high transport critical current density. Appl. Phys. Express 4, 063102 (2011).

77. Sun, D., Liu, Y. & Lin, C. Comparative study of upper critical field Hc2 and second magnetization peak Hsp in hole- and electron-doped BaFe2As2 superconductor. Phys. Rev. B 80, (2009).

78. Wang, X.-L. et al. Very strong intrinsic flux pinning and vortex avalanches in (Ba,K)Fe2As2 superconducting single crystals. Phys. Rev. B 82, (2010).

79. Yang, H., Luo, H., Wang, Z. & Wen, H.-H. Fishtail effect and the vortex phase diagram of single crystal Ba0.6K0.4Fe2As2. Appl. Phys. Lett. 93, 142506 (2008).

80. Scanlan, R. M., Malozemoff, A. P. & Larbalestier, D. C. Superconducting materials for large scale applications. Proc. IEEE 92, 1639–1654 (2004).

81. Braccini, V. et al. High-field superconductivity in alloyed MgB2 thin films. Phys. Rev. B 71, (2005).

82. Ozaki, T. et al. Fabrication of binary FeSe superconducting wires by novel diffusion process. at

111

83. Togano, K., Matsumoto, A. & Kumakura, H. Critical properties of a dense polycrystalline (Ba, K)Fe2As2superconductor prepared by a combined process of melting and deformation. Supercond. Sci. Technol. 23, 045009 (2010).

84. Wang, C. et al. Enhanced critical current properties in Ba0.6K0.4+xFe2As2 superconductor by overdoping of potassium. Appl. Phys. Lett. 98, 042508 (2011).

85. Durrell, J. et al. Critical Current of YBa2Cu3O7-δ Low-Angle Grain Boundaries. Phys. Rev. Lett. 90, (2003).

86. Song, X., Daniels, G., Feldmann, D. M., Gurevich, A. & Larbalestier, D. Electromagnetic, atomic structure and chemistry changes induced by Ca-doping of low-angle YBa2Cu3O7–δ grain boundaries. Nat. Mater. 4, 470–475 (2005).

87. Gao, Z., Togano, K., Matsumoto, A. & Kumakura, H. High transport J c in magnetic fields up to 28 T of stainless steel/Ag double sheathed Ba122 tapes fabricated by scalable rolling process. Supercond. Sci. Technol. 28, 012001 (2015).

88. Zhang, Z. et al. Effects of heating conditions on the microstructure and superconducting properties of Sr0.6K0.4Fe2As2. Supercond. Sci. Technol. 23, 065009 (2010).

89. Zhang, Z. et al. Heat treatment effects on the superconducting properties of Ag-doped SrKFeAs compounds. Sci. China Phys. Mech. Astron. 53, 1187–1193 (2010).

90. Kim, Y.-J., Weiss, J. D., Hellstrom, E. E., Larbalestier, D. C. & Seidman, D. N. Evidence for composition variations and impurity segregation at grain boundaries in high current- density polycrystalline K- and Co-doped BaFe2As2 superconductors. Appl. Phys. Lett. 105, 162604 (2014).

91. Pyon, S. et al. Effects of high-pressure annealing on critical current density in 122 type iron pnictide wires. Phys. C Supercond. 504, 69–72 (2014).

92. Gao, Z., Togano, K., Matsumoto, A. & Kumakura, H. Achievement of practical level critical current densities in Ba1−xKxFe2As2/Ag tapes by conventional cold mechanical deformation. Sci. Rep. 4, (2014).

93. Ma, Y. et al. Large transport critical currents and magneto-optical imaging of textured Sr1−xKxFe2As2 superconducting tapes. Supercond. Sci. Technol. 26, 035011 (2013).

112

94. Wang, L. et al. Textured Sr1−xKxFe2As2 superconducting tapes with high critical current density. Phys. C Supercond. 471, 1689–1691 (2011).

95. Scanlan, R. M., Fietz, W. A. & Koch, E. F. Flux pinning centers in superconducting Nb3Sn. J. Appl. Phys. 46, 2244–2249 (1975).

96. Matsushita, T., Kiuchi, M., Yamamoto, A., Shimoyama, J. & Kishio, K. Essential factors for the critical current density in superconducting MgB2: connectivity and flux pinning by grain boundaries. Supercond. Sci. Technol. 21, 015008 (2008).

97. Dew-Hughes, D. The role of grain boundaries in determining J c in high-field high-current superconductors. Philos. Mag. Part B 55, 459–479 (1987).

98. Haindl, S. et al. Thin film growth of Fe-based superconductors: from fundamental properties to functional devices. A comparative review. Rep. Prog. Phys. 77, 046502 (2014).

99. Iida, K. et al. Highly textured oxypnictide superconducting thin films on metal substrates. Appl. Phys. Lett. 105, 172602 (2014).

100. Sato, K. et al. High-Jc silver-sheathed Bi-based superconducting wires. IEEE Trans. Magn. 27, 1231–1238 (1991).

101. Hilgenkamp, H. & Mannhart, J. Superconducting and normal-state properties of YBa2Cu3O7−δ -bicrystal grain boundary junctions in thin films. Appl. Phys. Lett. 73, 265– 267 (1998).

102. Gurevich, A. & Pashitskii, E. A. Current transport through low-angle grain boundaries in high-temperature superconductors. Phys. Rev. B 57, 13878–13893 (1998).

103. Gurevich, A. et al. Flux Flow of Abrikosov-Josephson Vortices along Grain Boundaries in High-Temperature Superconductors. Phys. Rev. Lett. 88, (2002).

104. Chu, J.-H., Analytis, J. G., Kucharczyk, C. & Fisher, I. R. Determination of the phase diagram of the electron-doped superconductor $\text{Ba}{({\text{Fe}}_{1$- ${}x}{\text{Co}}_{x})}_{2}{\text{As}}_{2}$. Phys. Rev. B 79, 014506 (2009).

113

105. Hossain, M. A. et al. In situ doping control of the surface of high-temperature superconductors. Nat. Phys. 4, 527–531 (2008).

106. Aswathy, P. M., Anooja, J. B., Sarun, P. M. & Syamaprasad, U. An overview on iron based superconductors. Supercond. Sci. Technol. 23, 073001 (2010).

107. Campbell, A. M. & Evetts, J. E. Flux vortices and transport currents in type II superconductors. Adv. Phys. 21, 199–428 (1972).

108. Rose-Innes, A. C. & Rhoderick, E. H. Introduction to superconductivity. (Pergamon Press, 1978).

109. Matsushita, T., Otabe, E. S., Ni, B., Hikata, T. & Sato, K. History Effect of Critical Current Density and Weak Links in Superconducting Bi-Pb-Sr-Ca-Cu-O Tape Wires. TEION KOGAKU J. Cryog. Supercond. Soc. Jpn. 26, 384–388 (1991).

110. Müller, K.-H., Andrikidis, C. & Guo, Y. C. Irreversible critical current and the anomalous magnetic-moment peak in silver-sheathed (Pb, Bi)2Sr2Ca2Cu3O10+y tapes. Phys. Rev. B 55, 630–641 (1997).

111. Evetts, J. E. & Glowacki, B. A. Relation of critical current irreversibility to trapped flux and microstructure in polycrystalline YBa2Cu3O7. Cryogenics 28, 641–649 (1988).

112. Batista-Leyva, A. . et al. Hysteresis of the critical current density in YBCO, HBCCO and BSCCO superconducting polycrystals: a comparative study. Phys. C Supercond. 331, 57– 66 (2000).

113. Kunchur, M. N. & Askew, T. R. Hysteretic internal fields and critical currents in polycrystalline superconductors. J. Appl. Phys. 84, 6763 (1998).

114. McHenry, M., Maley, M. & Willis, J. Systematics of transport critical-current-density hysteresis in polycrystalline Y-Ba-Cu-O. Phys. Rev. B 40, 2666–2669 (1989).

115. Tarantini, C. et al. Artificial and self-assembled vortex-pinning centers in superconducting Ba(Fe1−xCox)2As2 thin films as a route to obtaining very high critical-current densities. Phys. Rev. B 86, (2012).

114

116. Matsumoto, A., Gao, Z., Togano, K. & Kumakura, H. Critical current density enhancement of uniaxially pressed Ag-sheathed (Sr,K)Fe2As2 superconducting wires. Supercond. Sci. Technol. 27, 025011 (2014).

117. Bottura, L. & Godeke, A. Superconducting Materials and Conductors: Fabrication and Limiting Parameters. Rev. Accel. Sci. Technol. 05, 25–50 (2012).

118. Miller, M. K., Russell, K. F., Thompson, K., Alvis, R. & Larson, D. J. Review of Atom Probe FIB-Based Specimen Preparation Methods. Microsc. Microanal. 13, 428–436 (2007).

119. Seidman, D. N. & Stiller, K. An Atom-Probe Tomography Primer. MRS Bull. 34, 717–724 (2009).

120. Seidman, D. N. Perspective: From field-ion microscopy of single atoms to atom-probe tomography: A journey: ‘Atom-probe tomography’. Rev. Sci. Instrum. 78, 030901 (2007).

121. Seidman, D. N. Three-Dimensional Atom-Probe Tomography: Advances and Applications. Annu. Rev. Mater. Res. 37, 127–158 (2007).

122. Larbalestier, D. C. et al. Experiments concerning the connective nature of superconductivity in YBa2Cu3O7. J. Appl. Phys. 62, 3308 (1987).

123. Sandim, M. J. R. et al. Grain boundary segregation in a bronze-route Nb3Sn superconducting wire studied by atom probe tomography. Supercond. Sci. Technol. 26, 055008 (2013).

124. Jo, Y. J. et al. High-field phase-diagram of Fe arsenide superconductors. Phys. C Supercond. 469, 566–574 (2009).

125. Tomita, M. & Murakami, M. High-temperature superconductor bulk magnets that can trap magnetic fields of over 17 tesla at 29 K. Nature 421, 517–520 (2003).

126. Durrell, J. H. et al. A trapped field of 17.6 T in melt-processed, bulk Gd-Ba-Cu-O reinforced with shrink-fit steel. Supercond. Sci. Technol. 27, 082001 (2014).

127. Larbalestier, D. C. et al. Strongly linked current flow in polycrystalline forms of the superconductor MgB2. Nature 410, 186–189 (2001).

115

128. Durrell, J. H. et al. A trapped field of 3 T in bulk MgB 2 fabricated by uniaxial hot pressing. Supercond. Sci. Technol. 25, 112002 (2012).

129. Yamamoto, A., Ishihara, A., Tomita, M. & Kishio, K. Permanent magnet with MgB2 bulk superconductor. Appl. Phys. Lett. 105, 032601 (2014).

130. Naito, T., Sasaki, T. & Fujishiro, H. Trapped magnetic field and vortex pinning properties of MgB 2 superconducting bulk fabricated by a capsule method. Supercond. Sci. Technol. 25, 095012 (2012).

131. Fuchs, G. et al. High trapped fields in bulk MgB2 prepared by hot-pressing of ball-milled precursor powder. Supercond. Sci. Technol. 26, 122002 (2013).

132. Evans, A. G. & Charles, E. A. Fracture Toughness Determinations by Indentation. J. Am. Ceram. Soc. 59, 371–372 (1976).

133. Dias, F. T. et al. Nanoscratch and indentation fracture toughness in superconductor Ba–Fe– As single crystals with lamellar structure. Tribol. Int. 79, 84–91 (2014).

134. Indrakanti, S. S. et al. Hot isostatic pressing of bulk magnesium diboride: Mechanical and superconducting properties. Philos. Mag. Lett. 81, 849–857 (2001).

135. Foerster, C. E. et al. Mechanical properties of Ag-doped top-seeded melt-grown YBCO pellets. Braz. J. Phys. 38, (2008).

136. Anstis, G. R., Chantikul, P., Lawn, B. R. & Marshall, D. B. A Critical Evaluation of Indentation Techniques for Measuring Fracture Toughness: I, Direct Crack Measurements. J. Am. Ceram. Soc. 64, 533–538 (1981).

137. Senkowicz, B. J. et al. Improved upper critical field in bulk-form magnesium diboride by mechanical alloying with carbon. Appl. Phys. Lett. 86, 202502 (2005).

138. Jiang, C. H., Dou, S. X., Cheng, Z. X. & Wang, X. L. Light carbon doping by oxygen-free paraffin wax to enhance the current density of MgB2 in the entire field regime. Supercond. Sci. Technol. 21, 065017 (2008).

139. Lin, H. et al. Hot pressing to enhance the transport Jc of Sr0.6K0.4Fe2As2 superconducting tapes. Sci. Rep. 4, 6944 (2014).

116

140. Gurevich, A. et al. Very high upper critical fields in MgB 2 produced by selective tuning of impurity scattering. Supercond. Sci. Technol. 17, 278–286 (2004).

141. Tarantini, C. et al. Development of very high Jc in Ba(Fe1-xCox)2As2 thin films grown on CaF2. Sci. Rep. 4, (2014).

142. Suryanarayana, C. Mechanical alloying and milling. Prog. Mater. Sci. 46, 1–184 (2001).

143. Lee, P. J. & Larbalestier, D. C. Microstructural factors important for the development of high critical current density Nb3Sn strand. Cryogenics 48, 283–292 (2008).

144. Bonney, L. A., Willis, T. C. & Larbalestier, D. C. Dependence of critical current density on microstructure in the SnMo6S8 Chevrel phase superconductor. J. Appl. Phys. 77, 6377 (1995).

145. Beek, C. J. van der et al. Electron irradiation of Co, Ni, and P-doped BaFe2As2–type iron- based superconductors. J. Phys. Conf. Ser. 449, 012023 (2013).

146. Pyon, S. et al. Enhancement of Critical Current Densities in (Ba,K)Fe2As2 by 320 MeV Au Irradiation in Single Crystals and by High-Pressure Sintering in Powder-in-Tube Wires. Appl. Phys. Express 6, 123101 (2013).

147. Taen, T. et al. Effects of irradiation-particle energy on critical current density in Co-doped BaFe2As2. Phys. C Supercond. 484, 62–65 (2013).

148. Eisterer, M. et al. Effects of disorder on the superconducting properties of BaFe1.8Co0.2As2 single crystals. Supercond. Sci. Technol. 22, 095011 (2009).

149. Zehetmayer, M. et al. Disorder induced effects on the critical current density of iron pnictide single crystals. Phys. C Supercond. 470, S452–S453 (2010).

150. Hecher, J. et al. The effect of neutron irradiation on interpreting the inter- and intragranular currents in fine-grain potassium doped Ba122 polycrystals with high global current density. (2013).

151. Zhang, X. et al. The influence of the sintering process on the superconducting property of Ba0.6K0.44Fe2As2.2. Supercond. Sci. Technol. 25, 035015 (2012).

117

152. Zhang, X. et al. Mechanism of enhancement of superconducting properties in a Ba1−xKxFe2As2 superconductor by Pb addition. Supercond. Sci. Technol. 25, 084024 (2012).

153. Maiorov, B. et al. Synergetic combination of different types of defect to optimize pinning landscape using BaZrO3-doped YBa2Cu3O7. Nat. Mater. 8, 398–404 (2009).

154. Miura, M. et al. Strongly enhanced flux pinning in one-step deposition of BaFe2(As0.66P0.33)2 superconductor films with uniformly dispersed BaZrO3 nanoparticles. Nat. Commun. 4, (2013).

155. Barone, A. Physics and applications of the Josephson effect. (Wiley, 1982).

118

BIOGRAPHICAL SKETCH

Jeremy D. Weiss

EDUCATION • Ph.D. in Materials Science and Engineering at Florida State University, May 2015 • IOP-ESAS 3rd Superconductivity Summer School, 2014 • National High Magnetic Field Laboratory User Summer School, 2014 • International Summer School on Superconductivity: Theory, Experiments, and Phenomena, 2013 • B.S. in Mechanical Engineering from Florida State University, 2010 • High school diploma from Douglas Anderson School of the Arts, 2005

HONORS AND AWARDS • Award for outstanding student presentation – Material Research Society Conference, 2014 • FSU Fellows Honor Society, 2014 • IEEE CSC Graduate Study Fellowship in Applied Superconductivity, 2013 • FSU Graduate Student Research and Creativity Award, 2013 • 1st place poster award – Applied Superconductivity Conference, 2012 • Publication featured on ceramics.org by American Ceramic Society: Ceramic Tech Today, 2012 • Golden Key International Honor Society, 2011

TEACHING EXPERIENCE • Mentor for National Science Foundation Research Experience for Undergraduates (NSF- REU) program, 2011, 2012, 2013, and 2014

• Substitute lecturer for EML3234 – Introduction to Material Science and Engineering. Fall 2011 and Spring 2012

119

RESEARCH EXPERIENCE

Graduate Research Assistant in the Applied Superconductivity Center at the National High Magnetic Field Laboratory, 2010-present.

Description: Research and development of advanced materials with an emphasis on the synthesis and characterization of electronic ceramics and processing of superconducting wires. Set up and managed a laboratory for handling hazardous materials. Supervised 1-3 employees. Contributed to grant proposals, authored scientific publications, and presented results at several conferences.

Supervisor: Prof. Eric Hellstrom

Projects: Investigation of Phase Relations and Reaction Pathways in Pnictide Superconductors and Understanding the Role of Grain Boundaries in Limiting current

Funding agency: NSF

Grant number: DMR-1006584 and DMR-1306785

Laboratory Research Associate in the Applied Superconductivity Center at the National High Magnetic Field Laboratory, 2007-2009.

Description: Worked on high temperature superconductor synthesis and material processing. Developed processing procedures, equipment, and experiments to handle and characterize iron-pnictide material safely. Microanalysis of high temperature superconductors.

Supervisor: Dr. Jianyi Jiang

Project: Fundamental questions about superconductivity in the Pnictides

Funding agency: AFRL, NSF, and the state of Florida

Grant number: FA9550-06-1-0474 and DMR-0084173

120

PROFESSIONAL SERVICE Service to the discipline • Session moderator, Applied Superconductivity Conference, 2014 • Student activities co-chair, Applied Superconductivity Conference, 2012 • Session moderator, Applied Superconductivity Conference, 2012 • Small Business Innovation Review (SBIR) selection process, Department of Energy, 2010 Occasional reviewer • Scientific Reports • IEEE Transactions on Applied Superconductivity • IEEE Applied Superconductivity Conference proceedings • Superconductor Science and Technology • Journal of Physics • U.S. Department of Energy Office of Science SBIR proposals

PROFESSIONAL PRESENTATIONS Weiss, J. D.; Collantes, Y.; Jiang J.; Tarantini, C.; Kametani, F.; Polyanskii, A. A.; Larbalestier,

D. C.; Hellstrom, E. E., Weak-links and grain boundary engineering in doped BaFe2As2 wires and bulks. (Talk) Applied Superconductivity Conference, Charlotte, NC. August 12 (2014)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Hainsey, B.; Larbalestier, D.C. and Hellstrom, E.E., Potassium Doped BaFe2As2 Wires and Bulks: Prospects for Applications at High Fields. (Invited Talk) Material Research Society Spring Meeting, San Francisco, CA. (2014)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Hainsey, B.; Larbalestier, D.C. and Hellstrom, E.E., Improvement and limitations of critical current densities in K-doped

ferropnictide BaFe2As2 bulks and wires. (Invited Talk) Electronic Materials and Applications Conference, Orlando, FL. January 23 (2014)

121

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Hainsey, B.; Larbalestier, D.C. and Hellstrom, E.E., Further improvement and limitations of critical current densities in K-doped ferropnictide BaFe2As2 bulks and multifilament round wires. (Talk) European Conference on Applied Superconductivity, Genova, Italy, September 15-19 (2013)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Hainsey, B.; Larbalestier, D.C.

and Hellstrom, E.E., High critical current density in (Ba0.6K0.4)Fe2As2 polycrystals and round wires with randomly oriented grains. (Poster) STEP 2013, Cargese, France, August 16 (2013)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Larbalestier, D.C. and

Hellstrom, E.E., Synthesis and Properties of High-Jc Bulk BaFe2As2 Superconductors. (Invited Talk) Electronic Materials and Applications Conference, Orlando, FL. (2013)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Larbalestier, D.C. and Hellstrom, E.E., Enhanced grain connectivity in K-doped ferropnictide Ba-122 bulks and wires with high transport critical current density. (Talk) SES American Physical Society meeting, Tallahassee, FL. (2012)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Larbalestier, D.C. and Hellstrom, E.E., Enhanced grain connectivity in K-doped Ba-122 wires with transport critical current density exceeding 0.1 MA/cm2. (Invited Talk) Applied Superconductivity Conference, Portland, OR. (2012)

Weiss, J.D.; Jiang, J.; Tarantini, C.; Kametani, F.; Polyanskii, A.; Larbalestier, D.C. and

Hellstrom, E.E., High critical current density in (Ba0.6K0.4)Fe2As2 polycrystals and round wires with randomly oriented grains. (Poster) Applied Superconductivity Conference, Portland, OR. (2012)

Weiss, J.D.; Jiang, J.; Hellstrom, E.E., Improvement of ferropnictide superconductors through novel synthesis techniques. (Poster) Material Research Society Spring Meeting, San Francisco, CA. (2011)

122

PUBLICATIONS, REFEREED ARTICLES Kim, Y; Weiss, J. D.; Hellstrom, E. E.; Larbalestier, D. C.; and Seidman, D. N. Evidence for composition variations and impurity segregation at grain boundaries in high current-density polycrystalline K- and Co-doped BaFe2As2 superconductors, Appl. Phys. Lett., 105, 162604 (2014)

Tarantini, C.; Kametani, F.; Lee, S.; Jiang, J.; Weiss, J.D.; Jaroszynski, J.; Hellstrom, E.E.; Eom,

C.B. and Larbalestier, D.C., Development of very high Jc in Ba(Fe1-xCox)2As2 thin films grown on CaF2, Scientific Reports, 4, 7305 (2014)

Lei, Q.Y.; Golalikhani, M.; Yang, D.Y; Withanage, W.K.; Rafti, A.; Qiu, J.; Hambe, M.; Bauer, E.D.; Ronning, F.; Jia, Q.X.; Weiss, J.D.; Hellstrom, E.E.; Wang, X.F.; Chen, X.H.; Williams,

F.; Yang, Q.; Temple, D. and Xi, X.X., Structural and transport properties of epitaxial Ba(Fe1- xCox)2As2 thin films on various substrates, Superconductor Science and Technology, 27 (11), 115010 (2014)

Nikolo, M.; Shi, X.; Jiang, J.; Weiss, J. D.; and Hellstrom, E. E., Magnetotransport Properties,

Thermally Activated FluxFlow, and Activation Energies in Ba(Fe0.95Ni0.05)2As2 and

Ba(Fe0.94Ni0.06)2As2 Superconductors, J. Supercond. Nov. Magn., 27, 1983-1990 (2014)

Perucchi, A.; Capitani, F.; Di Pietro, P.; Lupi, S.; Lee, S.; Kang, J.H.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E.; Eom, C.B.; Dore, P., Electrodynamics of superconducting pnictide superlattices. Appl. Phys. Lett., 104, 222601 (2014)

Nikolo, M.; Shi, X.; Choi, E.S; Jiang, J.; Weiss, J.D. and Hellstrom, E.E., Magneto-transport properties and thermally activated ﬂux ﬂow in Ba(Fe0.91Co0.09)2As2 superconductor, J. Supercond. Nov. Magn., (2014)

Weiss, J.D.; Jiang, J.; Polyanskii, A. A.; Hellstrom, E. E. Mechanochemical synthesis of pnictide

compounds and superconducting Ba0.6K0.4Fe2As2 bulks with high critical current density, Superconductor Science and Technology, 26, 074003 (2013)

123

Perucchi, A.; Baldassarre, L.; Joseph, B.; Lupi, S.; Lee, S.; Beom Eom, C.; Jiang, J.; Weiss, J. D.; Hellstrom, E. E. and Dore, P., Transmittance and reflectance measurements at terahertz frequencies on a superconducting BaFe1.84Co0.16As2 ultrathin film: an analysis of the optical gaps in the Co-doped BaFe2As2 pnictide, Eur. Phys. J. B, 86, 274 (2013)

Lee, S.; Tarantini, C.; Gao, P.; Jiang, J.; Weiss, J. D.; Kametani, F.; Folkman, C. M.; Zhang, Y.; Pan, X. Q.; Hellstrom, E. E.; Larbalestier, D. C.; and Eom, C. B. Artificially engineered superlattices of pnictide superconductor. Nature Materials, online (2013)

Tarantini, C.; Lee, S.; Kametani, F.; Jiang, J.; Weiss, J. D.; Jaroszynski, J.; Folkman, C. M.; Hellstrom, E. E.; Eom, C. B.; and Larbalestier, D. C. Artificial and self-assembled vortex- pinning centers in superconducting Ba(Fe1−xCox)2As2 thin films as a route to obtaining very high critical-current densities. Phys. Rev. B, 86, 214504 (2012)

Baldassarre, L.; Perucchi, A.; Postorino, P.; Lupi, S.; Marini, C.; Malavasi, L.; Jiang, J.; Weiss,

J.D.; Hellstrom, E.E.; Pallecchi, I. and Dore, P., Electrodynamics of BaFe2As2 from infrared measurements under pressure, Phys. Rev. B, 85, 174522 (2012)

Weiss, J.D.; Tarantini, C.; Jiang, J.; Kametani, F.; Polyanskii, A.A.; Larbalestier, D.C. and

Hellstrom, E.E., High intergrain critical current density in fine-grain (Ba0.6K0.4)Fe2As2 wires and bulks, Nature Materials, 11 (8), 682-685 (2012)

Celentano, G.; Marzi, G. De; Gaudio, S.; Augieri, A.; Galluzzi, V.; Mancini, A.; Rufoloni, A.; Vannozzi, A.; Corte, A. della; Gambardella, U.; Saggese, A.; Jiang, J.J.; Weiss, J. and Hellstrom,

E., The Effect of doping on the magnetic properties in Ba(Fe1-xCox)2As2 polycrystalline samples, IEEE Trans. Appl. Supercond., 21, 2874 (2011)

Khasanov, A.; Bhargava, S.C.; Stevens, J.G.; Jiang, J.; Weiss, J.D.; Hellstrom E.E. and Nath, A.,

Mössbauer studies of the superconducting cobalt/nickel-doped BaFe2As2. Whither go the injected electron (s)?, J. Phys. -Condens. Mat., 23, 202201(3pp) (2011)

124

Lee, S.; Jiang, J.; Weiss, J.D.; Bark, C.W.; Tarantini, C.; Biegalski, M.D.; Polyanskii, A.; Zhang, Y.; Nelson, C.T.; Pan, X.Q.; Hellstrom, E.E.; Larbalestier, D.C. and Eom, C.B., Dependence of epitaxial Ba(Fe1-xCox)2As2 thin films properties on SrTiO3 template thickness, IEEE Trans. Appl. Supercond., 21, 2882 (2011)

Yong, J.; Lee, S.; Jiang, J.; Bark, C.W.; Weiss, J.D.; Hellstrom, E.E.; Larbalestier, D.C.; Eom,

C.B. and Lemberger, T.R., Superfluid density measurements of Ba(CoxFe1−x)2As2 films near optimal doping, Phys. Rev. B, 83, 104510 (2011)

Zhang, Y.; Nelson, C.T.; Lee, S.; Jiang, J.; Bark, C.W.; Weiss, J.D.; Tarantini, C.; Folkman, C.M.; Baek, S.H.; Hellstrom, E.E.; Larbalestier, D.C.; Eom, C.B. and Pan, X.Q., Self-assembled

oxide nanopillars in epitaxial BaFe2As2 thin films for vortex pinning, Appl. Phys. Lett., 98, 042509 (2011)

Eisterer, M.; Zehetmayer, M.; Weber, H.W.; Jiang, J.; Weiss, J.D.; Yamamoto, A.; Hellstrom, E.E.; Larbalestier, D.C.; Zhigadlo, N.D. and Karpinski, J., Disorder effects and current percolation in FeAs-based superconductors, Superconductor Science and Technology, 23, 054006 (2010)

Gaudio, S.; Marzi, G.; Armenio, A.A.; Celentano, G.; Morici, L.; Corte, A.; Gambardella, U.; Jiang, J.; Hellstrom, E.E.; Weiss, J.D. and Larbalestier, D.C., Magnetic characterization of

Ba(Fe0.9Co0.1)2As2, Physica C, 470, s397 (2010)

Lee, S.; Jiang, J.; Nelson, C.; Bark, C.; Weiss, J.; Tarantini, C.; Jang, H.; Folkman, C.; Baek, S.; Polyanskii, A.; Abraimov, D.; Yamamoto, A.; Zhang, Y.; Pan, X.; Hellstrom, E.; Larbalestier, D.

and Eom, C., Template engineering of Co-doped BaFe2As2 single-crystal thin films, Nature Materials, 9, 397 (2010)

125

Mehta, M.; Sheet, G.; Dikin, D.A.; Lee, S.; Bark, C.W.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E.; Rzchowski, M.S.; Eom, C.B. and Chandrasekhar, V., Conductance asymmetry in point-contacts on epitaxial thin films of Ba(Fe0.92Co0.08)2As2, Appl. Phys. Lett., 97, 012503 (2010)

Perucchi, A.; Baldassarre, L.; Lupi, S.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E.; Lee, S.; Bark, C.W.; Eom, C.B.; Putti, M.; Pallecchi, I.; Marini, C. and Dore, P., Multi-gap superconductivity

in a BaFe1.84Co0.16As2 film from optical measurements at terahertz frequencies, Eur. Phys. J. B, 77, 25-30 (2010)

Putti, M.; Pallecchi, I.; Bellingeri, E.; Cimberle, M.R.; Tropeano, M.; Ferdeghini, C.; Palenzona, A.; Tarantini, C.; Yamamoto, A.; Jiang, J.; Jaroszynski, J.; Kametani, F.; Abraimov, D.; Polyanskii, A.; Weiss, J.D.; Hellstrom, E.E.; Gurevich, A.; Larbalestier, D.C.; Jin, R.; Sales, B.C.; Sefat, A.S.; McGuire, M.A.; Mandrus, D.; Cheng, P.; Jia, Y.; Wen, H.H.; Lee, S. and Eom, C.B., New Fe-based superconductors: properties relevant for applications, Superconductor Science and Technology, 23, 034003 (2010)

Sheet, G.; Mehta, M.; Dikin, D.A.; Lee, S.; Bark, C.W.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E.; Rzchowski, M.S.; Eom, C.B. and Chandrasekhar, V., Phase-incoherent superconducting pairs in the normal state of Ba(Fe1-xCox)2As2, Phys. Rev. Lett., 105 (16), 167003 (2010)

Tarantini, C.; Lee, S.; Zhang, Y.; Jiang, J.; Bark, C.W.; Weiss, J.D.; Polyanskii, A.; Nelson, C.T.; Jang, H.W.; Folkman, C.M.; Baek, S.H.; Pan, X.Q.; Gurevich, A.; Hellstrom, E.E.; Eom,

C.B. and Larbalestier, D.C., Strong vortex pinning in Co-doped BaFe2As2 single crystal thin films, Appl. Phys. Lett., 96, 142510 (2010)

Tortello, M.; Daghero, D.; Ummarino, G.A.; Stepanov, V.A.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E. and Gonnelli, R.S., Multigap superconductivity and strong electron-boson coupling in Fe- based superconductors: A point-contact andreev-reflection study of Ba(Fe1-xCox)2As2 single crystals, Phys. Rev. Lett., 105, 237002 (2010)

126

Valdés Aguilar, R.; Bilbro, L.S.; Lee, S.; Bark, C.W.; Jiang, J.; Weiss, J.D.; Hellstrom, E.E.; Larbalestier, D.C.; Eom, C.B. and Armitage, N.P., Pair-breaking effects and coherence peak in the terahertz conductivity of superconducting BaFe2−2xCo2xAs2 thin films, Phys. Rev. B, 82, 180514 (R) (2010)

Zehetmayer, M.; Eisterer, M.; Weber, H.W.; Jiang, J.; Weiss, J.D.; Yamamoto, A.; Polyanskii, A.A.; Hellstrom, E.E. and Larbalestier, D.C., Disorder induced effects on the critical current density of iron pnictide BaFe1.8Co0.2As2 single crystals, Physica C, 470, S452-S453 (2010)

Eisterer, M.; Weber, H.W.; Jiang, J.; Weiss, J.D.; Yamamoto, A.; Polyanksii, A.A.; Hellstrom,

E.E. and Larbalestier, D.C., Neutron irradiation of SmFeAsO1-xFx, Superconductor Science and Technology, 22, 065015 (2009)

Eisterer, M.; Zehetmayer, M.; Weber, H.W.; Jiang, J.; Weiss, J.D.; Yamamoto, A. and

Hellstrom, E.E., Effects of disorder on the superconducting properties of BaFe1.8Co0.2As2 single crystals, Superconductor Science and Technology, 22 (9), 095011 (2009)

Lee, S.; Jiang, J.; Weiss, J.D.; Folkman, C.M.; Bark, C.W.; Taratini, C.; Xu, A.; Abraimov, D.; Polyanskii, A.; Nelson, C.T.; Zhang, Y.; Baek, S.H.; Jang, H.W.; Yamamoto, A.; Kametani, F.; Pan, X.Q.; Hellstrom, E.E.; Gurevich, A.; Eom, C.B. and Larbalestier, D.C., Weak-link behavior

of grain boundaries in superconducting Ba(Fe1−xCox)2As2 bicrystals, Appl. Phys. Lett., 95, 212505 (2009)

127