Chapter 1

Complimentary Contributor Copy


MATERIALS SCIENCE AND TECHNOLOGIES

HIGH-TEMPERATURE SUPERCONDUCTORS OCCURRENCE, SYNTHESIS AND APPLICATIONS

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.


MATERIALS SCIENCE AND TECHNOLOGIES Additional books in this series can be found on Nova’s website under the Series tab.

Additional e-books in this series can be found on Nova’s website under the e-book tab.


MATERIALS SCIENCE AND TECHNOLOGIES

HIGH-TEMPERATURE SUPERCONDUCTORS OCCURRENCE, SYNTHESIS AND APPLICATIONS

MURALIDHAR MIRYALA AND

M. R. KOBLISCHKA EDITORS


Copyright © 2018 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. We have partnered with Copyright Clearance Center to make it easy for you to obtain permissions to reuse content from this publication. Simply navigate to this publication’s page on Nova’s website and locate the “Get Permission” button below the title description. This button is linked directly to the title’s permission page on copyright.com. Alternatively, you can visit copyright.com and search by title, ISBN, or ISSN. For further questions about using the service on copyright.com, please contact: Copyright Clearance Center Phone: +1-(978) 750-8400 Fax: +1-(978) 750-4470 E-mail: [email protected]. NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. Additional color graphics may be available in the e-book version of this book.

Library of Congress Cataloging-in-Publication Data ISBN: H%RRN

Published by Nova Science Publishers, Inc. † New York


CONTENTS Preface

ix

Chapter 1

Rebco Bulk Superconductors Doped with Sm or Al D. Volochová, V. Antal, S. Piovarči, V. Kavečanský and P. Diko

Chapter 2

Optimization of Flux Pinning and Growth Temperature for the Top-Seeded Infiltration Growth Processing of a Single Grain Bulk (Gd, Dy)BCO S. Pavan Kumar Naik, M. Murakami and M. Muralidhar

Chapter 3

The Applications of Superconducting Nanowire Network Fabrics Michael R. Koblischka, XianLin Zeng and Uwe Hartmann

Chapter 4

The Influence of an Initial Trapped Field on the Magnetic Shielding Performance of Bulk High-Temperature Superconducting Tubes L. Wéra, J. F. Fagnard, K. Hogan, B. Vanderheyden and P. Vanderbemden

Chapter 5

Experimental Investigation and Quantitative Analysis of the Normal-State Nernst Coefficient in Doped High-Temperature Superconductors of the YBa2Cu3Oy System Vitaliy E. Gasumyants and Olga A. Martynova

1

33 61

75

95

Chapter 6

Magnetic Alignment Techniques for HTSC Shigeru Horii and Jun-ichi Shimoyama

153

Chapter 7

Magnetization of Polycrystalline High-Tc Superconductors Denis Gokhfeld

181


vi Chapter 8

Contents Superconducting and Multiband Effects in FeSe with Ag Addition E. Nazarova, N. Balchev, K. Buchkov, K. Nenkov, D. Kovacheva, D. Gajda and G. Fuchs

Chapter 9

Magnetic Characterization of Bulk C-Added MgB2 Alex Wiederhold, Michael R. Koblischka, Miryala Muralidhar, Masato Murakami and Uwe Hartmann

Chapter 10

Recent Progress in Powder Densification of (Bi,Pb)-Sr-Ca-Cu-O Materials: Experimental Findings, Finite Element Simulations, and Practical Implications Lázaro Pérez-Acosta, Ernesto Govea-Alcaide, Renato de Figueiredo Jardim, Izabel Fernanda Machado, Fernando Rosales-Saíz, Sueli Hatsumi Masunaga, Jaime Eleicer Pérez-Fernández and Milton S. Torikachvili

Chapter 11

Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors A. Sedky

Chapter 12

Superconducting Motors and Generators Jean Lévêque, Kévin Berger and Bruno Douine

Chapter 13

A Review Article: Compact Magnetic Field Generators Containing HTS Bulk Magnets Cooled by Refrigerators and Their Feasible Applications Tetsuo Oka, Jun Ogawa, Satoshi Fukui, Takao Sato, Tomohito Nakano, Kazuya Yokoyama, Takashi Nakamura, Hiroyuki Fujishiro and Koshichi Noto

Chapter 14

Hybrid-Type Superconducting Magnetic Bearings for Rotating Machinery Jimin Xu, Zhi Li, Xiaoyang Yuan and Cuiping Zhang

Chapter 15

Recent Superconducting Applications in the Medical Field Santosh Miryala

Chapter 16

A Superconducting Magnetic Bearing Flywheel System Using a Multi-Surface Levitation Concept Selim Sivrioglu and Sinan Basaran


195

213

225

247 263

291

307 329

339

Contents Chapter 17

Strong Magnetic Field Generation by Superconducting Bulk Magnets Using Various Types of Refrigerators and Considering an Efficient Magnetization Kazuya Yokoyama and Tetsuo Oka

vii

369

About the Editors

387

Index

389



PREFACE Following the discovery of the oxide superconductors in 1986 by Bednorz and Müller, a number of new superconducting compounds was found exhibiting superconductivity above the liquid nitrogen temperature including rare-earth oxides and bismuth oxides, enabling cheaper cooling methods to be applied. The new class of superconducting materials featuring high critical temperatures (Tc above 77 K) and high second critical magnetic fields gave a new impetus to the research and development in superconductivity. Governments of many countries worldwide have encouraged scientists, engineers and industry to develop high temperature superconducting materials for practical use. Environmentally benign scenarios of high speed transport systems, energy saving by utilizing DC cables, medical equipment (e.g., magnetic resonance imaging (MRI) systems), a new class of magnetic drug delivery system (MDDS), nuclear magnetic resonance (NMR), and non-contact rotating machinery are expected. Nowadays, there is already a variety of high-Tc superconducting products available on the market, like single domain, batch processed bulks with diameters up to 140 mm, hundreds of kilometers of the first generation silver-sheathed Bi-2223 and Bi-2212 wires and tapes, and several kilometers of the second generation Y-123 tapes ('coated conductors') for winding coils of superconducting super-magnets or for constructing high-Tc super-cables for high energy transfer. The authors hope that, especially for young researchers, new upcoming engineers and students of superconductivity theory extension, the collected know-how in technology of superconducting thin films, wires, and bulks, and the new opportunities available for practical applications by the unique features of high-Tc materials will be very useful. The volume is designed to cover the recent achievements in occurrence, synthesis and application of high-Tc superconductors. The volume consists of a total of seventeen chapters, each of them defining in-depth the chapter subject and surveying recent developments in the field. The main objective of this volume is to summarize the recent advances in material science of high-Tc superconductors, including their properties,


x

Muralidhar Miryala and M. R. Koblischka

processing, and applications. New and challenging issues appear in this book, like superconducting large grain bulk RE-123, flux pinning, nanowire network fabrics and their applications, and a quantitative analysis on the normal-state Nernst coefficient. Furthermore, the book also covers recent developments on a variety of materials and the progress made, especially concerning the magnetic characterization of bulk C-doped MgB2, silver added bulk FeSe, and (BiPb)SrCaCuO systems, respectively. To show a full picture of the currently ongoing research efforts, the book covers large scale applications of bulk materials, including magnetic bearings, superconducting electric motors and their design layouts, hybrid-type superconducting magnetic bearings for rotating machinery, compact magnetic field generators, refrigerators, and recent developments in the application of superconducting super-magnets in the medical field. The authors would like to take this opportunity to express their sincere gratitude to all of the chapter contributors for their great endeavor in completing this book in time. They also wish to acknowledge Carra Feagaiga from NOVA Science Publisher for offering invaluable advice at every stage of editing this book. We would also like to extend our thanks to President Prof. M. Murakami-sensei, SIT for his constant support and encouragements. The team of authors and the editors sincerely hope that the presented ideas and information in this book will be helpful for interested readers, scientists, young researchers, bachelor and master students, and will encourage further development in the field. Muralidhar Miryala Editor Shibaura Institute of Technology Graduate School of Science and Engineering Superconductivity Research Laboratory (SRL) Tokyo, Japan Email: [email protected]

Michael R. Koblischka Editor Saarland University Institute of Experimental Physics Saarbrucken, Germany Email: [email protected]


In: High-Temperature Superconductors Editors: M. Miryala and M. R. Koblischka

ISBN: 978-1-53613-341-7 © 2018 Nova Science Publishers, Inc.

Chapter 1

REBCO BULK SUPERCONDUCTORS DOPED WITH Sm or Al D. Volochová*, V. Antal, S. Piovarči, V. Kavečanský and P. Diko Institute of Experimental Physics, Slovak Academy of Sciences, Košice, Slovak Republic

ABSTRACT The chapter is devoted to doped RE-Ba-Cu-O bulk superconductors (REBCO, RE = Y or rare earth element, here RE = Y or Gd). First, undoped YBCO single grain bulk superconductors with a nominal composition: 1 mol YBa 2Cu3O7-δ + 0.25 mol Y2O3 + 1 wt% CeO2 were prepared by the optimized top seeded melt growth (TSMG) process. Small single crystalline pieces cut from the SmBCO bulk crystal were used for seeding of epitaxial growth. Wavelength-dispersive spectrometry confirmed that prepared samples contain, besides samarium from the seed, also ytterbium from the substrate. The influence of this Sm and Yb contamination on superconducting properties of grown YBCO bulks is reported. It is shown that using a NdBCO seed and a combination of Y2O3 and Yb2O3 substrate leads to a high critical temperature (Tc(50%) = 91.54 K) and a sharp transition to the normal state (ΔTc = 0.45 K) of prepared undoped YBCO sample where the contamination is suppressed to a minimum. In the case of doped YBCO and GdBCO bulk superconductors, the effect of Sm and Al addition on the microstructure and superconducting properties of these materials has been studied. Nominal composition was enriched with different amounts of SmBa2Cu3Oy, Sm2O3 or Al2O3 (in the case of YBCO) and Gd2Ba6Al4O15 (in the case of GdBCO) powders with the aim to increase critical current density, Jc, especially in higher magnetic fields by introducing additional pinning centers. Single grain YBCO bulk superconductors with SmBa2Cu3Oy (Y123-Sm), Sm2O3 (Y123-SmO) or Al2O3 (Y123-Al) and single grain GdBCO bulk superconductors with Gd2Ba6Al4O15 (GdBCO-Al) powder addition were prepared by the optimized top seeded melt growth process. The influence of Sm and Al addition on the microstructure, * Corresponding Author Email: [email protected].


2

D. Volochová, V. Antal, S. Piovarči et al. critical transition temperature, Tc, critical current density, Jc, as well as maximum trapped magnetic field, Btmax, of prepared YBCO and GdBCO samples is reported. Additionally, in the case of Y123-Al samples, the possibilities of dopant redistribution by annealing in a controlled atmosphere are discussed.

Keywords: REBCO bulk superconductors, TSMG process, single grains, chemical pinning, microstructure, critical temperature, critical current density, peak effect, trapped magnetic field

1. INTRODUCTION Various practical applications of RE-Ba-Cu-O (REBCO, RE = Y or rare earth element) bulk superconductors are mainly based on their ability to trap large magnetic fields [1, 2]. It has already been shown that these materials can trap magnetic fields by order of magnitude higher than the best ferromagnets [3]. Recently, Durrell et al. have reported the largest trapped field to date of 17.6 T at 26 K, in a stack of two GdBCO superconducting bulks (Ø 25 mm) [4]. Generally, the magnitude of a trapped magnetic field in a bulk superconductor is proportional to the critical current density, Jc, and the length scale over which it flows (i.e., the diameter of a single grain sample). As a result, general processing aims of these materials consist in an enhancement of critical current density, Jc, of large single grain bulks. One of the mostly used melt processing techniques for production of YBCO (as well as REBCO generally) bulk superconductors is so-called top seeded melt growth (TSMG) process [5-7]. This process allows both production of large single grains as well as incorporation of effective pinning centers and consequent enhancement of critical current density, Jc, of these materials. TSMG proceess is based on the reversible peritectic reaction (TpY-123 ~ 1010°C) where the solid YBa2Cu3O7-δ (Y-123) phase decomposes to form a solid Y2BaCuO5 (Y-211) and a BaCuO-based liquid phase. In order to provide single heterogeneous nucleation, a small single crystal with lattice parameters similar to Y-123 is placed in the middle of the top surface of the pellet prior to melt processing. The Y-123 phase is formed by cooling a peritectically molten YBCO sample slowly through the peritectic temperature, Tp, (typically 0.1 - 1°C/h). During the growth, some Y-211 particles are trapped into the growing crystal and represent a basic structural component of such a YBCO melt processed superconductor. Since the critical current density, Jc, of these materials is indirectly proportional to the d211 (d211 - the size of the Y-211 particles) [8], they effectively contribute to flux line pinning when they are very fine. It is well known that a refinement of the Y-211 particles up to 0.5 - 1 μm may be reached by small additions of Pt or Ce-based compounds [9, 10].


REBCO Bulk Superconductors Doped with Sm or Al

3

The YBCO single grain bulk superconductors are commonly grown using either SmBCO (TpSm-123 ~ 1060°C) or NdBCO (TpNd-123 ~ 1070°C) seed crystals [11, 12] due to their higher peritectic temperature, or thin film seeds [13, 14]. They yield a processing window for the growth of Y-123, which becomes sufficiently large to decompose the YBCO pellet without melting the seed. Nevertheless, SmBCO seed dissolution has been reported [15-17], which leads to a diffusion of Sm into a sample volume. It was recently shown by Kim et al. [18] that a Y-211 block of proper thickness can be used as a diffusion barrier for samarium. Additional source of contamination may come from a substrate used for YBCO pellet. It has already been shown that melt processed YBCO samples grown on the Al2O3 substrate contain Al due to its dissolution in the melt [19]. Moreover, it is possible to introduce additional pinning centers, for example, by irradiation [20] or by chemical substitutions in a crystal lattice of a superconductor. In the latter case, different powder additions have been tested, for example Bi2O3 [21], NiFe2O4 [22], Pr6O11 [23], or Li2CO3 [24] leading in most cases to an increase of critical current density, Jc, and trapped magnetic field, Bt. In the first part of the chapter, we report on contamination of YBCO bulk superconductors by samarium from the seed and by ytterbium from the substrate. The influence of this contamination on the critical temperature, Tc, critical current density, Jc, and maximum trapped magnetic field, Btmax, of prepared YBCO samples is shown. Additionally, the microstructure and superconducting properties of YBCO bulk superconductors doped with Sm or Al and GdBCO bulk superconsuctors doped with Al are reported. The influence of Sm and Al addition on the microstructure, critical transition temperature, Tc, critical current density, Jc, as well as maximum trapped magnetic field, Btmax, of prepared YBCO and GdBCO samples is shown. Additionally, in the case of Y123-Al samples, the possibilities of dopants redistribution by annealing in a controlled atmosphere are discussed.

2. METHODS 2.1. Preparation of Undoped YBCO Bulk Superconductors Undoped YBCO single grain bulk superconductors (Y123) were fabricated by the top seeded melt growth (TSMG) process in air. 1 mol YBa2Cu3O7-δ, 0.25 mol Y2O3 and 1 wt% CeO2 were mixed in appropriate amounts for 30 min in a mixer and then intensively milled for 15 min in a mortar grinder. Mixture of powders was uniaxially pressed into the cylindrical pellets of 20 mm in diameter. The pressing was performed under a pressure of 60 kN. SmBCO or NdBCO single crystals were used as nucleation seeds in the TSMG process. To prevent pollution of the sample, Y-211 buffer layers (Ø5 mm, thickness 3 mm) were placed between the seed and the pellet [18] and the pellet


4

D. Volochová, V. Antal, S. Piovarči et al.

was placed on a thin layer of Yb2O3 + Y-123 powder mixture (60:40 wt%) as a substrate in order to prevent undesirable nucleation from the bottom (Figure 1). Later, three different kinds of substrates were tested: Yb 2O3 substrate, Y-211 and Yb2O3 substrate and Y2O3 and Yb2O3 substrate. The samples were treated in a chamber furnace with the time temperature regime optimized for high Y-123 crystal quality (Figure 2) [25].

Figure 1. The arrangements of the YBCO pellet prior the melt processing.

2.2. Preparation of YBCO Bulk Superconductors Doped with Sm Undoped YBCO single grain bulk superconductors (Y123) and samples with SmBa2Cu3Oy (Y123-Sm) or Sm2O3 (Y123-SmO) powder addition were fabricated by the TSMG process in air. 1mol YBa2Cu3O7, 0.25mol Y2O3 and 1wt% CeO2 powders with Z wt% of SmBa2Cu3Oy or Sm2O3 powder addition (Z = 0.05 - 10 wt% that was supposed to correspond to nominal x = 0.001, 0.0025, 0.005, 0.01, 0.02, 0.05 in Y(Ba1-xSmx)2Cu3O7-δ) were used in the composition of the samples of Y123-Sm or Y123-SmO, respectively. Precursor powders were mixed for 30 minutes in a mixer and then intensively milled for 15 minutes in a mortar grinder. Differential thermal analysis (DTA) was carried out on the precursor powders (mass of the each analysed sample: 100 mg) in flowing air at temperatures up to 1050°C with a heating rate 10°C/min prior to melt processing. Al2O3 crucibles were used for DTA measurements, and a baseline was applied for measurement correction. The mixed powders were uniaxially pressed into the cylindrical pellets of 20 mm in diameter. The samples were treated in a chamber furnace with the time temperature regime (Figure 2). NdBCO single crystals were used as nucleation seeds for the TSMG process. In this case, a combination of Y 2O3 and Yb2O3 substrates were used in order to keep the prepared samples without contamination.



5

Figure 2. Optimized time temperature regime of the TSMG process for preparation of undoped (Y123) and Sm doped (Y123-Sm and Y123-SmO) YBCO samples.

2.3. Preparation of YBCO Bulk Superconductors Doped with Al Al-doped Y123 single grain bulk superconductors (Y123-Al) were fabricated using the TSMG process in a chamber furnace with SmBCO seeds in an air atmosphere. As a nominal composition for the crystal growth we used a mixture of oxide powders YBa2Cu3O7-δ, Y2O3, and 0.5 wt% CeO2, each with a purity of at least 99.99%. In the case of doping by Al, the nominal composition was enriched with different amounts of Al 2O3, corresponding to different concentrations: x = 0.0025, 0.005, 0.02 and 0.05 of Al in YBa2(Cu1-xAlx)3O7-δ. As initial material YBa2Cu3O7-δ powder was taken with a maximum particle size of 30 µm. Oxide powders were mixed by milling using a mortar grinder with ZrO2 pestle for 15 min. After milling, the mixture of powders were sieved and uniaxially pressed into cylindrical pellets of 20 mm in diameter and 10 mm thickness. Undoped Y123 single grain bulk was prepared as a reference sample as well. Again, a combination of Y2O3 and Yb2O3 substrates were used in order to keep the prepared samples without contamination.

2.4. Preparation of GdBCO Bulk Superconductors Doped with Al Undoped and Al-doped single grain GdBCO bulk superconductors were fabricated by the TSMG process in a chamber furnace using SmBCO thin films evaporated onto MgO (100) substrates (SmBCO/MgO) as seed crystals. Commercially available powders


6


GdBa2Cu3O7-δ (Gd-123) and Gd2BaCuO5 (Gd-211) were mixed in a molar ratio Gd123:Gd-211 = 2:1 and 20 wt% Ag2O was added to improve the mechanical properties, 0.2 wt% Pt to refine the Gd211 particles. In the case of Al-doped GdBCO bulks (GdBCO-Al), Z wt% Gd2Ba6Al4O15 compound was added (Z = 0, 0.3, 0.5, 1.0, 2.0 and 5.5). All powders, in appropriate amounts, were mixed for 3 h in a milling machine incorporating an electrical mortar and pestle. Then the powder mixture was pressed uniaxially into pellets of 30 mm in diameter. The pellets were placed on ZrO2 rods. The typical temperature profile of the melt processing is shown in Figure 3.

Figure 3. Time temperature regime of the melt growth process for preparation of GdBCO bulk supraconductors undoped and doped with Al.

2.5. Structure and Microstructure Characterization The macrostructure of the sample surfaces was done by a stereo microscope. The microstructure characterization was examined by polarized light microscopy and by scanning electron microscopy (SEM) equipped with energy-dispersive x-ray spectrometry (EDS) as well as wavelength-dispersive x-ray spectrometry (WDS) on grinded and polished samples cut in the a/c-plane along the cylinder axis. The crystal lattice parameters of the Y123-Al samples were determined using x-ray diffraction analysis after both oxygen and argon annealing processes. Powders for x-ray analysis were prepared from as-grown bulks after TSMG process by their crashing and milling. Obtained powders were annealed and oxygenated in the same way as small samples for magnetisation measurements. Afterwards, x-ray diffraction scans were taken at room temperature with a conventional setup using CuKα radiation and an XPERT PRO



7

diffractometer. The lattices parameters of Y123 doped with Al (Y123-Al) and undoped (Y123) samples were determined from x-ray diffraction patterns using Rietveld refinement calculations.

2.6. Characterization of Superconducting Properties As grown single grain samples were oxygenated at 400°C for 200 h (Y123, Y123 doped with Sm) and at 410°C for 120 h (GdBCO and GdBCO-Al). After oxygenation the trapped field distribution of the bulk samples was mapped by the Hall probe scanning [26]. The samples were cooled to liquid nitrogen temperature in a magnetic field of 1.4 T applied parallel to the c-axis. The trapped field profiles were scanned 15 minutes after switching the external field off at a distance of 0.1 mm from the sample surface. Small specimens for oxygenation process, performed at 400°C for 200 h (Y123, Y123-Sm, Y123-SmO) and at 410°C for 120 h (GdBCO and GdBCO-Al) in a flowing oxygen atmosphere, and magnetization measurements were cut 0.5 mm below the top surface of the a-growth sector [27] of prepared bulks at a distance of approximately 2 mm from the seed. They had a shape of a slab with the dimensions approximately 1.5×1.5×0.5 mm3. The smallest dimension was parallel to the c-axis of the crystal. In the case of Y123-Al samples, the annealing processes of Al-doped and undoped reference Y123 samples were performed in pure oxygen and argon gases using two annealing processes. During the first annealing process, one series of the samples was slowly heated to 800°C in flowing oxygen, annealed for two hours, slowly cooled down to 400°C and oxygenated there for 240 h. The second series of the samples was annealed also at 800°C for two hours, this time in flowing argon and slowly cooled down to room temperature. After annealing in argon, the samples were oxygenated in a tubular furnace at 400°C for 240 h. The superconducting properties of the samples were measured using a vibrating sample magnetometer (VSM). The field dependence of the magnetic moment was measured for all samples at 77 K with a magnetic field of up to 6 T at a constant sweep rate of 0.25 T/min. During the magnetisation measurements, the applied magnetic field was parallel to the c-axis of the crystal. The critical current densities, Jc, were calculated from the magnetic hysteresis loops using the Bean model [28] for rectangular samples, Jc(B) = {Δm(B)/V}{2/[b(1-b/3a)]}, where Δm is the difference in magnetic moments between the ascending and descending branches of the magnetic hysteresis loops, and V is the sample volume, a×b×c. The critical transition temperatures, Tc, were determined from the magnetic transition curves at 50% of the low temperature magnetization. The magnetizations were measured in an applied external magnetic field of 2 mT after zero field cooling. The transition widths, ΔTc, were determined from the same curve subtracting the 90% and 10% of the low temperature magnetization.


8


3. RESULTS AND DISCUSSION 3.1. Contamination of YBCO Bulk Superconductors The top surface macrograph of YBCO sample prepared by applying the optimized TSMG process in air [25] and using a SmBCO seed crystal is shown in Figure 4. Fully grown single grain sample can be seen in this figure.

Figure 4. Top surface macrograph of YBCO single grain sample grown using the optimized TSMG process.

Figure 5. Magnetization measurements of YBCO sample prepared by the optimized TSMG process using a SmBCO seed crystal.

Magnetization measurements revealed relatively low critical temperature, Tc, of the prepared YBCO sample – Tc(50%) = 84.7 K (Figure 5(a)). Additionally, peak effect was observed in the field dependence of critical current density, Jc, (Figure 5(b)). These observations are usually typical for chemically doped samples where impurity atoms substitute atoms in the crystal lattice of superconductor, suppress superconductivity locally and can act as additional pinning centers [29, 30]. Therefore we analyzed the concentration of possible elements which could pollute the sample during melt processing by WDS line microanalysis. It can be seen in Figure 6 that Sm from the seed and Yb from the substrate were identified in concentrations 0.1 wt% and 0.25 wt%, respectively, practically in the whole sample volume.



9

Figure 6. WDS microanalysis of the YBCO sample prepared by the optimized TSMG process using a SmBCO seed and Yb2O3 + Y-123 substrate. R denotes the radius of the sample and h is its height.

Contamination of YBCO samples by Sm has already been reported [18, 31, 32]. Dissolution of used SmBCO seed during TSMG melt processing of Y-123 was observed [15]. It was supposed that at high temperatures the seed/melt interface is thermodynamically unstable which results in partial or full dissolution of the used seed during melt processing [15, 16]. In order to minimize this Sm pollution, we performed several experiments. First, we decreased the holding time on Tmax to 1 h. It was observed by optical microscopy that 1 h is time long enough to melt the whole sample [33]. This led to the increase of Tc(50%) from 84.7 K (Figure 5(a)) to 87.5 K (Figure 7(a)). Nevertheless, the critical temperature was found to be still low and pronounced peak effect can be seen in Jc(B) dependence (Figure 7(b)). We obtained similar results even we decreased Tmax (Tmax = 1030°C or 1020°C) and used Y-211 buffer layer (Y-211 BL, Ø 5 mm, thickness 3 mm) [18] (Figure 7). It can be supposed now that low Tc and peak effect are caused by Yb pollution as well which is present in the whole sample volume as was shown by WDS analysis (Figure 6(b)). In order to minimize contamination of YBCO samples by Yb, different kinds of substrates were tested. In these experiments, NdBCO single crystals were used as a nucleation seed. It can be seen in Figure 8 that using a Y2O3-Yb2O3 substrate leads to high Tc and sharp transition to normal state (Tc(50%) = 91.5 K, Tconset = 92.04 K, ΔTc = 0.45 K). In this case the pronounced peak effect was suppressed, and we suppose that the contamination of prepared YBCO sample was minimal. The trapped magnetic field profile of this YBCO sample is shown in Figure 9(b). It can be compared with a trapped magnetic field profile of the YBCO sample contaminated by Yb from the substrate (Figure 9(a)). The increase of the maximum trapped magnetic field, Btmax, can be clearly recognized.


10


Figure 7. Magnetization measurements of YBCO samples prepared at different Tmax with dwell time 1 h. In some cases Y-211 buffer layer (Y-211 BL) acting as a diffusion barrier for Sm was placed between the YBCO pellet and a SmBCO seed.

Figure 8. Magnetization measurements of YBCO samples grown using different kinds of substrates. NdBCO single crystals were used as seeds.

Figure 9. Trapped magnetic field profiles of YBCO samples prepared using SmBCO seeds and different kinds of substrates - Yb2O3+Y-123 substrate (a) and Y2O3 - Yb2O3 substrate (b).



11

3.2. YBCO Bulk Superconductors Doped with Sm Figure 2 shows the time temperature regime of the TSMG process for the preparation of undoped YBCO single grain bulk superconductors (Y123) and the samples with Sm addition (Y123-Sm, Y123-SmO). It can be seen that two different temperatures of isothermal dwell, Tis, were used. Tis = 998°C in the case of the samples Y123, Y123-Sm (x = 0.001 - 0.01) and the samples Y123-SmO (x = 0.001 - 0.05) and Tis' = 1006°C in the case of samples Y123-Sm (x = 0.02 and 0.05). The requirement of the optimization of the TSMG process for the samples with higher Sm-123 powder additions is related to the complexity of the fabrication process of these materials in the form of single grains. The series of our papers as well as others publications on growth of YBCO bulk superconductors [25, 34, 35] clearly demonstrate this complexity. A single crystal growth from the seed requires finding a temperature interval above the temperature of selfnucleation, TSN, of the Y-123 phase in the field where heterogeneous nucleation from the seed occurs. This temperature interval is usually very narrow, several degrees for the YBCO system with Y-211 addition, up to 10°C for the YBCO system with Y2O3 addition in nominal composition. Therefore, DTA measurements were performed on precursor powders containing 1mol YBa2Cu3O7 + 0.25mol Y2O3 + 1wt% CeO2 (nominal composition) and different Sm-123 or Sm2O3 powder additions (Figure 10). DTA involves heating or cooling a studied sample and an inert reference (empty Al 2O3 crucible) under identical conditions, while recording any temperature difference between the sample and reference. DTA signals for all powder compositions are plotted against temperature in Figure 10. Generally, changes in the sample, either endothermic or exothermic, can be detected relative to the inert reference. In Figure 10, two endothermic peaks in each studied composition can be clearly recognized. The first one is related to the reaction where Y2O3 reacts with a part of Y-123 phase to form Y-211 and a small amount of CuO: (ssr) YBa2Cu3O7-δ + 1.5Y2O3 → 2 Y2BaCuO5 + CuO + 0.5 O2

Tssr < 940°C

(1)

Moreover, CuO formed by reaction (1) reacts with a part of Y-123 phase according to reaction: (p1) a YBa2Cu3O7-δ + b CuO ↔ c Y2BaCuO5 + d L(p1) + e O2

Tp1= 940°C

(2)

The second, usually called high temperature endothermic peak, is related to the reversible peritectic reaction: (m1)

YBa2Cu3O7-δ ↔ a Y2BaCuO5 + b L(m1) + c O2 TpY-123 = 1010°C in air (3)


12


A shift of the high temperature endothermic peak is clearly visible on DTA traces in Figure 10(a), especially in the case of the Y123-Sm, x = 0.02 and 0.05. As a result, these samples required the growth at higher temperatures employing the temperature of isothermal dwell, Tis' = 1006°C. In the case of the samples with Sm2O3 powder addition, the influence of the addition on high temperature endothermic peak position was not significant (Figure 10(b)), and all the samples were grown using Tis = 998°C.

Figure 10. DTA measurements on precursor powders of nominal composition and different Sm-123 (a) or Sm2O3 (b) powder additions. A shift of high temperature endothermic peak can be clearly seen in the case of higher concentrations of Sm-123 powder addition (Figure 10(a)).

Figure 11 shows top surface macrographs of undoped Y123 (a), Y123-Sm, x = 0.001 (b) and x = 0.01 (c), Y123-SmO, x = 0.001 (d) and x = 0.05 (e) single grain samples prepared by the optimized TSMG process. No spontaneous parasitic grain nucleation was observed showing the suitability of the used TSMG process. Typically melt processed YBCO bulk superconductor consists of five growth sectors (GSs): 4 a-GSs and one c-GS divided by growth sector boundaries (GSBs). The cross section of the Y123-Sm, x = 0.05 sample in Figure 12 shows c-GS as well as a-GS divided by GSBs. The concentration of Y-211 particles in the growth sectors is not homogeneous, as can be seen in this figure. Their concentration is lower in the c-GS than in a-GS making the growth sector boundary well visible under polarised light. Different distribution of Y-211 particles could be explained by pushing and trapping theory [36] which proposed that when a sample solidifies at a certain growth rate, solid particles present in the melt that are smaller than critical are not trapped in the solid. Critical particle size is smaller in the a-GS than in the c-GS.



13

Note: a-growth sector: a-GS, growth sector boundaries: GSBs. Figure 11. Top surface macrographs of Y123 (a), Y123-Sm, x = 0.001 (b), Y123-Sm, x = 0.01 (c), Y123-SmO, x = 0.001 (d) and Y123-SmO, x = 0.05 (e) single grain samples.

Note: a-growth sector: a-GS, c-growth sector: c-GS, growth sector boundaries: GSBs. Figure 12. Cross section of the Y123-Sm, x = 0.05 sample.

In order to determine the influence of the Sm-123 or Sm2O3 additions on the microstructure of YBCO bulk superconductors, the samples without any addition and with the highest concentration of Sm were studied using polarised light microscopy. Because of the mentioned Y-211 particle distribution inhomogeneity, it is important to note that Figure 13 compares the microstructure of the samples Y123 (Figure 13(a),(d)), Y123-Sm, x = 0.05 (Figure 13(b),(e)) and Y123-SmO, x = 0.05 (Figure 13(c),(f)) taken from the beginning of the c-GS and a-GS of each sample. It can be seen that Sm2O3 addition leads to a higher amount of smaller Y-211 particles in both c-growth sector as well as a-growth sector of the Y123-SmO, x = 0.05 sample (Figure 13 (c),(f)). The authors suppose that a part of Sm2O3 powder reacts with Y-123 phase to form (Y,Sm)-211 particles (according to reaction (1)) what may be related to such a microstructure. Therefore, the Sm2O3 addition seems to be useful from the point of the size, d211, and volume fraction, V211, of the Y-211 particles. Since the value of Jc at low magnetic fields is proportional to V211/d211 [8], an enhancement of Jc(0T, 77K) of the Y123-SmO samples is expected.


14


Figure 13. Microstructure of the samples Y123 (c-GS (a), a-GS (d)), Y123-Sm, x = 0.05 (c-GS (b), aGS (e)) and Y123-SmO, x = 0.05 (c-GS (c), a-GS (f)).

In order to determine the influence of Sm addition (in the form of Sm-123 as well as Sm2O3 powder) on the superconducting properties, a series of magnetization measurements was performed on prepared Y123-Sm and Y123-SmO samples using a vibrating sample magnetometer. It should be noted that all the specimens were cut from the equivalent position within the bulk (as described in the Part 2) and they were all oxygenated at same time under the same thermal conditions in order to keep the comparison meaningful. The results of the magnetization measurements are summarized in Figure 14 and Figure 15. The critical temperatures, Tc, of the samples with different Sm concentrations were not found to be significantly influenced by Sm-123 and Sm2O3 addition when compared to the undoped Y123 sample (x = 0, Tc = 90.9 K). Moreover, Sm-123 as well as Sm2O3 addition in the studied range does not influence the width of superconducting transition, ΔTc, and ΔTc is around 0.5 K for all samples (Figure 14). Microstructure analysis performed by polarized light microscope revealed that Sm123 addition leads to a higher amount of slightly coarser Y-211 particles (Figure 13(b),(e)) that is related to lower critical current densities (Jc(0T,77K) ~ 6 × 104 A/cm2) of the YBCO samples with Sm-123 addition in low magnetic fields when compared to the undoped YBCO sample (Jc(0T,77K) = 6.6 × 104 A/cm2) (Figure 15(a)) [8]. Additionally, an increase of critical current density, Jc(0 T, 77 K), of the samples with Sm2O3 addition (all concentrations) compared to the undoped YBCO sample was observed (Figure 15(b)). Moreover, an enhancement of the critical current density, Jc, in higher magnetic fields (2-3 T) may suggest the presence of some additional pinning centres created by Sm substitutions (Figure 15(a),(b)).



15

Figure 14. Tc and ΔTc values as a function of Sm concentration, x, for the YBCO samples with Sm-123 addition (Y123-Sm) (a) and Sm2O3 addition (Y123-SmO) (b).

Figure 15. Field dependences of the critical current density at 77 K and H // c of the samples without any addition (Y123) and samples with Sm-123 addition (Y123-Sm) (a) and Sm2O3 addition (Y123SmO) (b).

Finally, the maximum trapped magnetic field, Btmax, was measured in the prepared single grain Y123-Sm bulk superconductors. In spite of relatively low enhancement of Jc, all the additional samples with Sm-123 possess a trapped magnetic field higher than that of the undoped YBCO sample (Table 1). A maximum trapped magnetic field, Btmax, of 564 mT at 77 K in Y123-Sm, x = 0.0025 sample (Ø17.2 mm) was 43% higher than that of the undoped YBCO sample (Figure 16). Table 1. Dependence of the maximum trapped field, Btmax, of the Y123-Sm single grain samples on the concentration of Sm, x Concentration, x Btmax (mT)

0 395

0.001 462

0.0025 564

0.005 513

0.01 500


0.02 504

16


Figure 16. Profile of trapped magnetic field at 77 K in the undoped Y123 sample, Btmax = 395 mT (a), and in the Y123-Sm sample, x = 0.0025, Btmax = 564 mT (b).

3.3. YBCO Bulk Superconductors Doped with Al 3.3.1. Superconducting Properties of Al-Doped Y123 High critical current density is an important parameter of the superconductors for their practical application. The critical current density (Jc) of Y123 can be substantially increased by introducing (creating) non-superconducting regions acting as effective pinning centres. Chemical substitution of atoms in the Y123 crystal lattice is one of the most widely used method for creating pinning centres. In this section we show how the critical current density in Y123 is enhanced after chemical substitution of Cu atoms in the CuO chains by Al atoms and influence of this substitution on the superconducting transition temperature. It was pointed out in experimental section that different nominal Al concentrations from x = 0.0025 to 0.05 were used for chemical substitution, and additional high-temperature annealing processes in oxygen and argon were applied for deeper investigation of possible different Al arrangement within the samples. The critical current density for both oxygen and argon annealing processes are presented in Figure 17(a) and (b), respectively. The critical current density at the lowest Al concentration (x = 0.0025) for annealing in oxygen is higher from 0.5 to 2 T in comparison to the undoped Y123 sample due to the appearance of the clear peak effect (Figure 17(a)). Thus, the critical current density improves by Al doping. The critical current density at the highest Al concentration (x = 0.05) is not presented for the oxygen annealed samples, because this value is negligibly small. From the Jc results of the oxygen annealed samples it is clear that the lowest Al concentration is the most optimal in our Al concentration range to reach necessary amount of pinning centres with the size of approximately two coherence lengths (about 6 nm at 77 K) when the flux pinning is the most effective [37]. When the dopant concentration increases, the



17

mean distance between the pinning centres becomes shorter and the locally disordered regions overlap, which should lead to a decrease in the critical current density with increasing nominal Al concentration [38].

Figure 17. Dependence of the critical current densities, Jc, on applied magnetic field, B, for YBa2(Cu1xAlx)3O7-δ after high-temperature annealing in oxygen (a) and argon (b).

Figure 18. The superconducting transition temperatures (a) and transition widths (b) as a function of nominal Al concentration for YBa2(Cu1−xAlx)3O7-δ samples after annealing in oxygen and argon.

The superconducting transition temperatures, Tc, and transition widths, ∆Tc, for oxygen and argon annealed samples are presented in Figure 18(a) and (b), respectively. Tc monotonously decreases with increasing nominal Al concentration (Figure 18(a)), and ∆Tc (Figure 18(b)) is wider for the samples annealed in oxygen compared to annealing in argon. Decreasing Tc and increasing ∆Tc for Al-doped Y123 samples were ascribed to the microscopic inhomogeneity of Al arrangement resolved in the Y123 phase [39]. Additional heat treatment of Al-doped Y123 samples in flowing argon at 800°C for 2 hours led to significant changes in both the superconducting transition temperature


18


(Figure 18(a),(b)) and the pinning behaviour (Figure 17(b)). Tc is above 90 K and remains high with increasing Al concentration up to x = 0.02, similar to the undoped Y123 reference sample, and ∆Tc are almost similar and not influenced by Al atoms in the whole concentration range. The clear peak effect on Jc(B) curves is observed between argon annealed samples only at the highest Al concentration (x = 0.05) as well as the decrease of Tc. In the case of annealing in argon, a decrease of Tc, a presence of the peak effect up to nominal Al concentration x = 0.02 and unchanged ∆Tc (Figure 18(b)) are not observed because of the possibility of creating Al clusters during the heat treatment at 800°C. The distance between the disturbed regions where Tc might be locally suppressed increases by clustering Al atoms, and Tc retains its original value in between [40]. Thus Al doped samples behave as the undoped Y123 reference samples. It should be noted that the substitution by the trivalent atoms in the CuO chains can have an influence on the total carrier density [41, 42] and the oxygen content [43] of the samples which reflect on the critical transition temperature or pinning behaviour. The tendency to cluster can be explained by the need of trivalent atoms, such as Al, Fe or Co, substituted in the CuO chains to share oxygen atoms in order to increase their coordination number [44, 45]. The decrease of Tc (Figure 18(a)) in the argon annealed sample at the highest Al concentration (x = 0.05) may be caused by the impossibility to cluster all Al atoms due to the high Al concentration. Appearance of the peak effect (Figure 17(b)) at the highest Al concentration (x = 0.05) indicates that sufficient concentration of the pinning centres was created and the distance between them was suitable for effective flux pinning. It is evident from Jc and Tc results that high-temperature annealing in different atmospheres can influence the superconducting properties of Al-doped Y123 bulk superconductors. It should be noted that different arrangement of the substituent atoms in the Y123 crystal lattice was also observed in Ni-doped and Li-doped Y123 samples, for example. But there are some differences in Ni and Li distribution within the samples in comparison to Al-doped Y123 since Ni and Li can substitute Cu atoms in both the CuO chains and the CuO2 planes. After high-temperature annealing of Ni-doped Y123 in oxygen, Ni atoms substitute Cu atoms in both the CuO chains and the CuO 2 planes, whereas after annealing at lower oxygen partial pressure, the migration of Ni atoms from the CuO chains to the CuO2 planes takes place [46, 47], and finally, this arrangement of Ni influences the superconducting properties of Y123. In the case of Li-doped Y123 samples, powder neutron diffraction measurements [48, 49] showed that during chemical substitution, the arrangement of Li atoms in Y123 depends on the fabrication process and thermo-chemical treatments. It was shown that in Li-doped Y123 samples synthesised in oxygen, Li atoms substitute Cu atoms in the CuO2 planes, whereas in the samples grown in air or annealed in argon, Li atoms can substitute Cu atoms in both the CuO 2 planes and the CuO chains. However, not more than a quarter of Li substitutes Cu atoms in the CuO



19

chains after the preparation of Li-doped Y123 in air or annealed in argon as the rest of Li prefers substitution of Cu atoms in the CuO2 planes. Moreover, different distribution of Ni in comparison to Li-doped Y123 is caused by different valence of Ni and Li and further maintaining its favourede coordination number.

3.3.2. Microstructure Analysis of Al-Doped Y123 Previous microstructure analysis of Al-doped Y123 samples has shown that Al does not form any secondary phases with Y, Ba, Cu or Ce in our nominal Al concentration range [50]. More interesting results were obtained in the investigation of the twinning structure of the samples. The oxygen annealed Al-doped Y123 samples shows that the spacing between the twin boundaries decreases with increasing Al content, and in such a way the transition from the orthorhombic to the tetragonal (pseudotetragonal) state takes place. The twin structures of Al-doped Y123 samples at Al concentration of x = 0.005 and x = 0.05 after annealing in oxygen are shown in Figure 19(a) and (b), respectively. The similar decrease of the spacing between the twin boundaries with increasing concentration of Al was found in Al-doped Y123 polycrystalline ceramic samples [39, 51]. When the undoped Y123 sample has the orthorhombic structure, Cu atoms in the CuO chains have four-fold oxygen coordination. If the trivalent Al atoms substitute for Cu atoms in the CuO chains, they prefer five-fold oxygen coordination. Al atoms can get five-fold oxygen coordination in the case if the so-called extra oxygen atoms are added to adjacent (1/2,0,0) sites at the a-axis of the crystal lattice between the CuO chains [52], as in the case of Co and Fe doped samples [53, 54]. The extra oxygen atoms are adopted by Al atoms which now acquire a coordination number equal to five. Extra oxygen atoms are associated with the formation of the twinning structures in Y123, and they are the initial nucleation centres for new twin boundaries. The number of extra oxygen atoms increases with Al concentration in Y123, and they are randomly distributed within the sample. With increasing oxygen content of Y123 during the oxygenation process, the twin boundaries are spreading within the sample from the initial nucleation centers. In such a way Al atoms can influence the twinning process. Correspondingly, there are more small twins in the oxygen annealed samples at the highest Al concentration (x = 0.05) and the spacing between the twin boundaries is reduced (Figure 19(b)). The twin structures of YBa2(Cu1-xAlx)3O7-δ samples annealed in argon at 800°C with x = 0.005 and 0.05 are presented in Figure 20(a) and (b), respectively. No significant changes in the twin structures are observed for both samples, especially at the highest Al concentration (x = 0.05). This demonstrates that Al atoms after annealing the samples in argon do not have a strong influence on twinning as in the case of oxygen annealing.


20


Figure 19. Micrographs of YBa2(Cu1-xAlx)3O7-δ in polarized light showing the twinning structure (bright and dark lamellas) for x = 0.005 (a) and x = 0.05 (b) after annealing in oxygen. Darker parts in the micrographs are Y2BaCuO5 non-superconducting particles.

Figure 20. Micrographs of YBa2(Cu1-xAlx)3O7-δ in polarized light showing the twin structure (bright and dark lamellas) for x = 0.005 (a) and x = 0.05 (b) for samples annealed in argon. Darker parts in the micrographs are Y2BaCuO5 non-superconducting particles.

Figure 21. Schematic illustration of the possible distribution of Al atoms and influence of extra oxygen atoms on the twin structure within Al-doped Y123 sample after annealing in oxygen (a) and clustered Al atoms after annealing in argon (b). The red circles in (a) and (b) correspond to Figure 22(a) and (b), respectively.



21

Figure 22. Arrangement of Al atoms in the Y123 crystal lattice in random case (a) and after clustering (b). Crystal lattices (a) and (b) correspond to the red circles in Figure 21(a) and (b), respectively.

The CuO chains could be partially formed during high-temperature annealing in oxygen at 800°C. Al atoms get enough oxygen for creating the favoured five-fold oxygen coordination and do not need to form clusters in order to share oxygen atoms, as it is necessary in the case of annealing in argon. If Al atoms are clustered, the extra oxygen atoms are localized mainly in the clusters. The extra oxygen atoms located in Al clusters do not have so strong an influence on the twin structure, and consequently clustered Al atoms do not suppress Tc and do not induce the peak effect on Jc(B). An evidence confirming Al clustering during annealing in argon can be the wider spacing between the twin boundaries (Figure 20(b)) at the highest Al concentration (x = 0.05). Schematic illustrations of the possible random distributions of Al atoms after annealing in oxygen and clustered Al atoms after annealing in argon are shown in Figure 21(a) and (b), respectively. Possible arrangement of the compound atoms in the crystal lattice of Al-doped Y123 samples after annealing in oxygen and argon are presented in Figure 22(a) and (b), respectively. There are some examples to show six-fold coordinated Al atoms in the clusters as well. In the clusters, two Al atoms share one extra oxygen atom at the a-axis of the crystal lattice (Figure 22(b)).

3.3.3. X-Ray Diffraction Analysis of Al-Doped Y123 The dependence of the crystal lattice parameters of Y123 phase on nominal Al concentration obtained using the X-ray diffraction analysis is presented in Figure 23(a). As it was expected, lattice parameters a and b grow closer with increasing Al concentration (Figure 23(a)). It is obvious that the tendency of the phase transition from the orthorhombic to the tetragonal (pseudotetragonal) state occurs faster for the samples


22


annealed in oxygen than in the case of the argon annealed samples. The same tendency for the orthorhombic to the tetragonal transition is observed for the twin structure of oxygen annealed samples (Figure 19(b)) with the highest Al concentration (x = 0.05). The differences between the crystal lattice parameters of oxygen and argon annealed Aldoped Y123 samples are better presented by the orthorhombicity, in Figure 23(b). The orthorhombicity decreases slower with the increasing Al concentration in the argon annealed samples than after annealing in oxygen. This confirms that Al atoms form clusters during high-temperature annealing in argon.

Figure 23. The crystal lattice parameters of Y123 phase for YBa2(Cu1-xAlx)3O7-δ (a) and variation of the orthorhombicity, Φ, after annealing in oxygen and argon (b) as a function of nominal Al concentration.

3.4. GdBCO Bulk Superconductors Doped with Al 3.4.1. Macro- and Microstructure of Samples The single grain GdBCO samples with Z = 0, 0.3, 0.5, 1.0 and 2.0 wt% of Gd2Ba6Al4O15 powder addition were successfully grown, while the multicrystalline sample was obtained for the addition Z = 5.5 wt% of this powder. The typical examples of the top surface of the single grain samples are presented in Figure 24.

Figure 24. Example of top surface macrographs of prepared GdBCO + Z wt% Gd2Ba6Al4O15 single grain bulk samples. Z = 0 (a), 0.5 (b).



23

In the micrographs taken at higher magnifications in polarized light from the polished cross sections of the single grain samples along the a/c-plane, Gd-211 particles and Ag particles were observed (Figure 25(a)). Their size and distribution did not depend on Gd2Ba6Al4O15 powder additions. The Gd-211 particle size is below one micrometer which means that their growth during processing was effectively retarded by Pt addition. Only in the single grain sample with the highest Al concentration (2.0 wt% of Gd 2Ba6Al4O15) as well as in the multicrystalline sample with 5.5 wt% addition of Gd 2Ba6Al4O15 powder some other secondary phase was detected, besides Ag and Gd-211 phases (Figure 25(b)). This phase is present in the form of plates and appears as dark particles with elongated shape on the sample cross section. It has similar morphology as Pt,Al-based phase, (Ba3Y1)Pt(CuAl)2O9+δ which was observed in the YBCO bulks with Pt addition grown on Al2O3 rods [19]. EDS analysis of this phase (Figure 26) confirmed that it is composed of Gd, Ba, Cu, Al, Ag and Pt (Table 2). As the signal for EDS analysis comes also from the matrix surrounding the phase, it is not possible to estimate its elemental stoichiometry. We may suppose that in our case some Ag also joined this phase. Observation of this phase only in the samples with 2.0 wt% and 5.5 wt% of Gd 2Ba6Al4O15 means that in the case of lower Gd2Ba6Al4O15 additions, Al is dissolved in the Gd-123 lattice.

Figure 25. Gd-211 particles (dark) and Ag particles (bright) in the sample with 0.3 wt% Gd 2Ba6Al4O15 addition (oxygenated sample) (a) and Pt,Al-based phase with rectangular cross section in the sample with 2.0 wt% of Gd2Ba6Al4O15 addition (as grown sample) (b).

Table 2. Typical cation composition of Pt,Al-type secondary phase estimated by EDS analysis Element At %

Al 5.84

Pt 1.73

Ag 2.66

Ba 20.6

Gd 19.92


Cu 48.90

24


Figure 26. SEM micrograph of the microstructure of the sample with 2.0 wt% of Gd2Ba6Al4O15 addition. Besides Ag and Gd-211 globular particles, elongated Pt,Al-based phase can also be seen (a). EDS spectrum of the Pt,Al-based phase marked by arrow (b).

3.4.2. Trapped Field The results of trapped field measurements are shown in Figure 27. The maximum trapped field value, Btmax, was obtained for the addition of 0.3 wt% of Gd 2Ba6Al4O15 powder and then decreases with Gd2Ba6Al4O15 powder addition. The addition of 0.3 wt% Gd2Ba6Al4O15 corresponds to GdBa2(Cu0.9975Al0.0025)3O7-δ if we suppose that all Al substituted Cu in Gd-123 phase. The decreasing trapped field at higher Al concentrations in the Gd-123 phase can be caused by two facts. The first is decreasing transition temperature with increased Al doping [50] seen in Figure 28. The critical transition temperature, Tc, of Al-doped GdBCO samples decreases monotonously with increasing Gd 2Ba6Al4O15 powder addition, Z. The second reason can be related to the optimum Al concentration for singleatom pinning [38, 55] which is close to the lowest Al doping level used in this study.

Figure 27. Dependence of maximum trapped field, Btmax, on Gd2Ba6Al4O15 addition, Z (a) and trapped field profile of the sample with 0.3 wt% Gd2Ba6Al4O15 addition (b).



25

Figure 28. Dependence of the critical transition temperature, Tc, on Z wt% of Gd2Ba6Al4O15 powder addition.

CONCLUSION Undoped YBCO single grain bulk superconductors were prepared by the optimized TSMG process in air. The critical temperatures were found to be lower that for the Y123 phase (Tc ~ 92 K), even Y-211 buffer layer or NdBCO seed were used. Decrease of Tmax did not have significant influence on Tc. Peak effect in the field dependences of the critical current density at 77 K was observed. WDS line microanalysis confirmed that prepared samples contain, besides 0.1 wt% Sm from the seed, also 0.25 wt% Yb from the substrate. Decrease of dwell time at Tmax = 1040°C to 1h, use of a NdBCO seed and a combination of Y2O3 and Yb2O3 substrate led to a high critical temperature with a sharp transition to the normal state (Tc(50%) = 91.54 K, Tconset = 92.04 K, ΔTc = 0.45 K) of prepared YBCO sample. In this case pronounced peak effect was suppressed and a higher maximum trapped magnetic field was reached. Therefore, the contamination of the samples was supposed to be suppressed to a minimum. YBCO bulk superconductors with Sm addition were prepared by the optimised TSMG process in the form of single grains using Sm-123 or Sm2O3 powder additions. Microstructure analysis revealed that Sm2O3 powder addition leads to a higher amount of smaller Y-211 particles which is related to higher critical current density, Jc, in low magnetic fields of the Y123-SmO samples. Additionally, an enhancement of Jc in higher magnetic fields in Y123-Sm (x = 0.001, 0.0025, 0.05) and Y123-SmO (all samples) suggests the presence of additional pinning centres created by Sm substitutions. The critical temperatures, Tc, as well as the width of superconducting transition, ΔTc, were not significantly influenced by Sm addition. From the point of YBCO bulk superconductors with Sm addition, Sm2O3 powder addition seems to be more suitable, mainly because of


26


its positive influence on Y-211 particles refinement and consequent enhancement of critical current density, Jc. In spite of relatively low enhancement of Jc in the case of Y123-Sm samples, higher values (up to 43%) of the maximum trapped magnetic fields, Btmax, measured in all the Y123-Sm samples represent a positive influence of Sm addition on the properties of YBCO bulk superconductors. Moreover, we have studied YBCO single grain bulk superconductors doped with Al. Obtained results confirm that aluminium is suitable for chemical substitution of Cu atoms in Y123 single grain bulk superconductors prepared by the TSMG process. The created point defects act as effective pinning centres, and the critical current densities can be significantly improved. The chemical substitutions by Al was studied before but effective flux pinning was not observed because the samples with very high doping concentrations were examined. Also, we used CeO2 addition in our Al-doped Y123 samples for the refinement of Y211 particles whereas up to now other studies on Al-doped Y123, the samples were fabricated using Pt addition. The lowest nominal Al concentration (x = 0.0025) is considered as optimal doping for oxygen annealed samples. In this case the critical transition temperature was not essentially affected, but the critical current density was improved by the presence of a pronounced peak effect on Jc(B) curve at 77 K in comparison to the undoped Y123 sample. High-temperature annealing of Al-doped YBCO samples in argon causes significant changes in both Tc and pinning behaviour. The critical transition temperature does not decrease with the increase of Al content, and no peak effect is observed up to x = 0.02. This is associated to the clustering of Al atoms during annealing in argon. After clustering a greater part of the samples retains their original value and is undisturbed by Al substitution. Experimental results show that Al atoms are mobile during hightemperature annealing and, depending on the high-temperature annealing atmosphere, different Al arrangements in Al-doped Y123 samples can be obtained. Moreover, aluminium does not form any secondary phases with CeO2 technological addition whereas with platinum it can form, for example, (Ba3Y1)Pt(CuAl)2O9+δ. Finally, we have successfully prepared single grain GdBCO bulk superconductors with the addition of Gd2Ba6Al4O15 powder by the TSMG process in air. Microstructure analysis confirmed that up to 1.0 wt% addition of this powder, Al-based phase is not formed and Al is supposed to substitute Cu in GdBa2Cu3O7-δ phase. In the sample with 2.0 wt % addition of Gd2Ba6Al4O15 powder, Pt,Al-based phase is formed indicating that the solubility limit of Al in the system was exceeded. Al doping increased maximum trapped field value, Btmax, at 0.3 wt % Gd2Ba6Al4O15 addition and reached 0.9 T at 77 K. The critical transition temperature, Tc, of prepared Al-doped GdBCO samples decreases monotonously with increasing Gd2Ba6Al4O15 powder addition.



27

ACKNOWLEDGMENTS This work was accomplished within the framework of the projects: New Materials and Technologies for Energetic (ITMS 26220220061), Research and Development of Second Generation YBCO Bulk Superconductors (ITMS 26220220041), APVV No. 0330-12, VEGA No. 2/0121/16.

REFERENCES [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

Fuchs, G., Schätzle, P., Krabbes, G., Gruß, S., Verges, P., Müller, K. H., Fink, J., Schultz, L. 2000. “Trapped magnetic fields larger than 14 T in bulk YBa2Cu3O7−x.” Appl. Phys. Lett. 76: 2107. doi: http://dx.doi.org/10.1063/1.126278. Werfel, F. N., Floegel-Delor, U., Rothfeld, R., Riedel, T., Goebel, B., Wippich, D., Schirrmeister, P. 2012. “Superconductor bearings, flywheels and transportation.” Supercond. Sci. Technol. 25: 014007. doi: https://doi.org/10.1088/0953-2048/25/ 1/014007. Tomita, M., Murakami, M. 2003. “High-temperature superconductor bulk magnets that can trap magnetic fields of over 17 tesla at 29 K.” Nature 421: 517-520. doi:10.1038/nature01350. Durrell, J. H., Dennis, A. R., Jaroszynski, J., Ainslie, M. D., Palmer, K. G. B., Shi, Y. H., Campbell, A. M., Hull, J., Strasik, M., Hellstrom, E. E., Cardwell, D. A.. 2014. “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. doi: https://doi.org/10. 1088/0953-2048/27/8/082001. Morita, M., Takebayashi, S., Tanaka, M., Kimura, K., Miyamoto, K., Sawano, K. 1991. “Quench and Melt Growth (QMG) Process for Large Bulk Superconductor Fabrication.” Adv. in Supercond. III: 733-736. doi: https://doi.org/10.1007/978-4431-68141-0_162. Izumi, T., Nakamura, Y., Shiohara, Y. 1992. “Diffusion solidification model on Ysystem superconductors.” J. Mater. Res. 7: 1621-1628. doi: https://doi.org/ 10.1557/JMR.1992.1621. Murakami, M., Yoo, S. I., Higuchi, T., Sakai, N., Weltz, J., Koshizuka, N., Tanaka, S. 1994. “Flux Pinning in Melt-Grown NdBa2Cu3Oy and SmBa2Cu3Oy Superconductors.” Jpn. J. Appl. Phys. 33: L715-L717. doi: https://doi.org/10.1143/JJAP.33.L715. Murakami, M., Yamaguchi, K., Fujimoto, H., Nakamura, N., Taguchi, T., Koshizuka, N., Tanaka, S. 1992. “Flux pinning by non-superconducting inclusions


28

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

D. Volochová, V. Antal, S. Piovarči et al. in melt-processed YBaCuO superconductors.” Cryogenics 32: 930-935. doi: https://doi.org/10.1016/0011-2275(92)90002-R. Ogawa, N., Hirabayashi, I., Tanaka, S. 1991. “Preparation of a high-Jc YBCO bulk superconductor by the platinum doped melt growth method.” Physica C 177: 101105. doi: https://doi.org/10.1016/0921-4534(91)90304-H. Kim, C. J., Lai, S. H., McGinn, P. J. 1994. “Morphology of Y2BaCuO5 and segregation of second phase particles in melt-textured Y-Ba-Cu oxides with/without BaCeO3 addition.” Mater. Lett. 19: 185-191. doi: https://doi.org/10.1016/0167577X(94)90066-3. Cardwell, D. A., Hari Babu, N. 2006. “Processing and properties of single grain (RE)–Ba–Cu–O bulk superconductors.” Physica C 445-448: 1-7. doi: https://doi.org/10.1016/ j.physc.2006.03.065. Iida, K., Babu, N. H., Shi, Y., Matthews, L. S., Miyazaki, T., Sakai, N., Murakami, M., Cardwell, D. A. 2007. “Phase stability of MgO-doped Nd–Ba–Cu–O seed crystals for cold-seeded, OCMG processing.” Physica C 463-465: 340-343. doi: https://doi.org/ 10.1016/j.physc.2007.01.037. Yao, X., Nomura, K., Nakamura, Y., Izumi, T., Shiohara, Y. 2002. “Growth mechanism of high peritectic temperature Nd 1+xBa2−xCu3O7−δ thick film on low peritectic temperature YBa2Cu3O7−δ seed film by liquid phase epitaxy.” J. Cryst. Growth 234: 611-615. doi: https://doi.org/10.1016/S0022-0248(01)01760-2. Li, T. Y., Wang, C. L., Sun, L. J., Yan, S. B., Cheng, L., Yao, X., Xiong, J., Tao, B. W., Feng, J. Q., Xu, X. Y., Li, C. S., Cardwell, D. A. 2010. “Multiseeded melt growth of bulk Y–Ba–Cu–O using thin film seeds.” J. Appl. Phys. 108: 023914. doi: http://dx.doi.org/ 10.1063/1.3465301. Jee, Y. A., Hong, G. W., Kim, C. J., Sung, T. H. 1998. “Dissolution of SmBa2Cu3O7-y seed crystals during top-seeded melt growth of YBa2Cu3O7-y.” Supercond. Sci. Technol. 11: 650-658. doi: https://doi.org/10.1088/09532048/11/7/007. Jee, Y. A., Hong, G. W., Sung, T. H., Kim, C. J. 1999. “Effect of resolidification nature of SmBa2Cu3O7−y seeds on the growth mode of YBa2Cu3O7−y.” Physica C 314: 211-218. doi: https://doi.org/10.1016/S0921-4534(99)00021-0. Jee, Y. A., Hong, G. W., Kim, C. J., Sung, T. H. 1999. “Dissolution and resolidification of SmBa/sub 2/Cu/sub 3/O/sub 7-y/ seed during top seeded melt texturing of YBa/sub 2/Cu/sub 3/O/sub 7-y/.” IEEE Trans. Appl. Supercond. 9: 2097-2100. doi: 10.1109/77.784880. Kim, C. J., Lee, J. H., Park, S. D., Jun, B. H., Han, S. C., Han, Y. H. 2011. “Y2BaCuO5 buffer block as a diffusion barrier for samarium in top seeded melt growth processed YBa2Cu3O7 − y superconductors using a SmBa2Cu3O7 − d seed.” Supercond. Sci. Technol. 24: 015008. doi: https://doi.org/10.1088/09532048/24/1/015008.



29

[19] Diko, P., Stöver, G., Krabbes, G. 2002. “Morphology and composition changes of the Pt-based secondary phase in melt-grown YBCO caused by Al pollution.” Supercond. Sci. Technol. 15: 217-221. doi: https://doi.org/10.1088/09532048/15/2/306. [20] Gonzalez-Arrabal, R., Eisterer, M., Weber, H. W., Fuchs, G., Verges, P., Krabbes, G. 2002. “Very high trapped fields in neutron irradiated and reinforced YBa2Cu3O7−δ melt-textured superconductors.” Appl. Phys. Lett. 81: 868-870. doi: http://dx.doi.org/ 10.1063/1.1497705. [21] Yang, W. M., Wang, M. 2013. “New method for introducing nanometer flux pinning centers into single domain YBCO bulk superconductors.” Physica C 493: 128-131. doi: https://doi.org/10.1016/j.physc.2013.03.045. [22] Li, G. Z., Dong, L., Deng, X. Y. 2016. “Significant Improvement of Superconducting Properties in Nano-NiFe2O4-Doped Y–Ba–Cu–O Single-Grain Superconductor.” J. Am. Ceram. Soc. 99: 388-391. doi: 10.1111/jace.14035. [23] Xu, X. Q., Cai, Y. Q., Yang, C. X., Yao, X., Xu, S., Kortyka, A., Puzniak, R. 2009. “Effect of Pr doping on the growth and superconducting properties of (Y1−xPrx)Ba2Cu3O7−δ.” Supercond. Sci. Technol. 22: 015001. doi: https://doi.org/10.1088/0953-2048/22/1/015001. [24] Shlyk, L., Krabbes, G., Fuchs, G., Nenkov, K., Verges, P. 2003. “Melt-textured YBa2Cu3O7 based material doped with Li: high trapped fields and pinning.” Physica C 392-396: 540. doi: https://doi.org/10.1016/S0921-4534(03)00741-X. [25] Volochová, D., Diko, P., Radušovská, M., Antal, V., Piovarči, S., Zmorayová, K., Šefčiková, M. 2012. “Growth of Y123 bulk crystals in Y1.5Ba2Cu3Ox system with CeO2 addition.” J. Cryst. Growth 353: 31-34. doi: https://doi.org/10.1016/j.jcrysgro. 2012.04.029. [26] Murakami, M., 2001. “Measurements of trapped-flux density for bulk hightemperature superconductors.” Physica C 357-360: 751-754. doi: https://doi.org/10.1016/S0921-4534(01)00380-X. [27] Diko, P. 2000. “Growth-related microstructure of melt-grown REBa2Cu3Oy bulk superconductors.” Supercond. Sci. Technol. 13: 1202-1213. doi: https://doi.org/10. 1088/0953-2048/13/8/317. [28] Bean, C. P. 1962. “Magnetization of hard superconductors.” Phys. Rev. Lett. 8: 250-253. doi: https://doi.org/10.1103/PhysRevLett.8.250. [29] Shlyk, L., Krabbes, G., Fuchs, G., Stöver, G., Gruss, S., Nenkov, K. 2002. “Pinning behavior and magnetic relaxation in melt-processed YBCO doped with Li, Ni and Pd.” Physica C 377: 437-444. doi: https://doi.org/10.1016/S0921-4534(01)01298-9. [30] Antal, V., Kanuchova, M., Sefcikova, M., Kovac, J., Diko, P., Eisterer, M., Horhager, N., Zehetmayer, M., Weber, H. W., Chaud, X. 2009. “Flux pinning in Al doped TSMG YBCO bulk superconductors.” Supercond. Sci. Technol. 22: 105001. doi: https://doi.org/ 10.1088/0953-2048/22/10/105001.


30


[31] Dewhurst, C. D., Lo, W., Shi, Y. H., Cardwell, D. A. 1998. “Homogeneity of superconducting properties in SmBa2Cu3O7−δ-seeded melt processed YBCO.” Mater. Sci. Eng. B 53: 169-173. doi: https://doi.org/10.1016/S0921-5107(98) 80009-3. [32] Chen, P. W., Chen, I. G., Chen, S. Y., Wu, M. K. 2011. “The peak effect in bulk Y– Ba–Cu–O superconductor with CeO2 doping by the infiltration growth method.” Supercond. Sci. Technol. 24: 085021. doi: https://doi.org/10.1088/09532048/24/8/085021. [33] Diko, P., Radušovská, M. Unpublished result. [34] Volochová, D., Diko, P., Antal, V., Radušovská, M., Piovarči, S. 2012. “Influence of Y2O3 and CeO2 additions on growth of YBCO bulk superconductors.” J. of Cryst. Growth 356: 75-80. doi: https://doi.org/10.1016/j.jcrysgro.2012.07.021. [35] Wang, W., Peng, B., Chen, Y., Guo, L., Cui, X., Rao, Q., Yao, X. 2014. “Effective Approach to Prepare Well c-Axis-Oriented YBCO Crystal by Top-Seeded MeltGrowth.” Cryst. Growth Des. 14: 2302-2306. doi: 10.1021/cg5000028. [36] Endo, A., Chauchan, H. S., Egi, T., Shiohara, Y. 1996. “Macrosegregation of Y2Ba1Cu1O5 particles in Y1Ba2Cu3O7−δ crystals grown by an undercooling method.” J. Mater. Res. 11: 795-803. doi: https://doi.org/10.1557/JMR.1996.0096. [37] Takezawa, N., Fukushima, K. 1997. “Optimal size of an insulating inclusion acting as a pinning center for magnetic flux in superconductors: calculation of pinning force.” Physica C 290: 31-37. doi: https://doi.org/10.1016/S0921-4534(97)01574-8. [38] Ishii, Y., Yamazaki, Y., Nakashima, T., Ogino, H., Shimoyama, J., Horii, S., Kishio, K. 2008. “Chemical (Sr,Co)-doping effect on critical current density for Dy123 melt-solidified bulks.” Mater. Sci. Engin. B 151: 69-73. doi: https://doi.org/10.1016/j.mseb.2008.02.005. [39] Siegrist, T., Schneemeyer, L. F., Waszczak, J. V., Singh, N. P., Opila, R. L., Batlogg, B., Rupp, L. W., Murphy, D. W. 1987. “Aluminum substitution in Ba2YCu3O7.” Phys. Rev. B 36: 8365-8368. doi: https://doi.org/10.1103/Phys RevB.36.8365. [40] Brecht, E., Schmahl, W. W., Miehe, G., Rodewald, M., Fuess, H., Andersen, N. H., Hanßmann, J., Wolf, Th. 1996. “Thermal treatment of YBa2Cu3-xAlxO6+δ single crystals in different atmospheres and neutron-diffraction study of excess oxygen pinned by the A1 substituents.” Physica C 265: 53-66. doi: https://doi.org/10.1016/ 0921-4534(96)00255-9. [41] Ming, L. 1994. “Band Structures of Al-doped Superconductors YBa2Cu3-xAlxO7+δ.” Int. J. Quantum Chem. 50: 233-242. doi: 10.1002/qua.560500402. [42] Clayhold, J., Hagen, S., Wang, Z. Z., Ong, N. P., Tarascon, J. M., Barboux, P. 1989. “Chain-site versus plane-site Cu substitution in YBa2Cu3-xMxO7 (M=Co,Ni): Hall and thermopower studies.” Phys. Rev. B 39: 777-780. doi: https://doi.org/10.1103/ PhysRevB.39.777.



31

[43] Jorgensen, J. D., Veal, B. W., Paulikas, A. P., Nowicki, L. J., Crabtree, G. W., Claus, H., Kwok, W. K. 1990. “Structural properties of oxygen-deficient YBa2Cu3O7-δ.” Phys. Rev. B 41: 1863-1877. doi: https://doi.org/10.1103/Phys RevB.41.1863. [44] Maury, F., Mirebeau, I., Hodges, J. A., Bourges, P., Sidis, Y., Forget, A. 2004. “Long- and short-range magnetic order, charge transfer, and superconductivity in YBa2Cu3-xCoxO7.” Phys. Rev. B 69: 094506-10. doi: https://doi.org/10.1103/ PhysRevB.69.094506. [45] Renevier, H., Hodeau, J. L., Marezio, M., Santoro, A. 1994. “Electron- and powder neutron-diffraction studies of YBa2(Cu1-yCoy)3O6+x with 0.05 Jc//c [9]. Furthermore, the increase in the misorientation angle between two RE123 grains leads to serious reduction of the intergrain Jc even for the coriented RE123 grains [10, 11]. These two important experimental works indicate that simultaneous formation of c-axis grain-orientation and the alignment of grains along the a- and b-axis directions, being parallel to the CuO2 planes, bi- or tri-axial grainorientation, is required to improve cuprate superconductor materials to a practical stage. As described above, the grain orientation is a key issue for the improvement of Jc in the cuprate superconducting materials. In the case of RE123, thin film growth [12] and melt-solidification [13] processes based on an epitaxial growth technique have been standard processes to form bi-axial oriented microstructures. On the other hand, in the case of (Bi,Pb)2223, mechanical uni-axial orientation has been used as a production process of practical Ag-sheathed superconducting tapes [14]. From the viewpoint of ideal HTS wires, one can recognize that the disadvantages of RE123 coated conductors and Ag-sheathed (Bi,Pb)2223 tapes are thin HTS layers and lack of bi-axial orientation of HTS materials, respectively. Development of HTS materials with both large crosssectional area and high degrees of bi-axial orientation is important for production of the “ideal” HTS wires. Magnetic alignment techniques are focused on in this chapter. Magnetic alignment shows two advantageous points as grain-orientation methods of HTS materials. One is that recent developments in superconducting magnets brought us a 10 T class of static magnetic fields in room temperature bore without using liquid helium [15, 16]. As shown in Equation (1), such high magnetic fields can produce orientation energy comparable to thermal energy at room temperature, even for feeble magnetic


156

Shigeru Horii and Jun-ichi Shimoyama

(paramagnetic and diamagnetic) substances [17]. The other is that magnetic alignment using MRF [18, 19] is a tri-axial (or bi-axial) grain-orientation technique. The principle of three different magnetic grain-orientation techniques including alignment of the first easy axis, the hard easy axis and all three magnetization axes is quantitatively described in the next section. Furthermore, the magnetic alignment techniques are applicable by combination with colloidal processes, such as slip-casting, electrophoretic deposition, sheet-casting and so on. One can easily imagine that the combination of magnetic alignment using MRF and sheet-casting leads to the production of “ideal” HTS tapes with a large cross-sectional area and high degrees of tri-axial orientation in principle. However, in order to apply this combined process for the fabrication of thick and triaxially oriented superconducting ceramics, fundamental researches on bahaviors of powders of HTSC under various magnetic fields are required as functions of the type of elements (or RE), magnetic field conditions, and mean diameters of the powders. Particularly in RE123, a twin microstructure containing two different domains in which these c-axes are shared and these a- and b-axes are orthogonal each other, is formed in its grain. One should take into account the possibility that in-plane magnetic anisotropy in RE123 grain with the twin microstructure is reduced or disappears and the tri-axial orientation degrees are reduced. In this chapter, the principle of three different magnetic alignment techniques is described at first. In order to understand magnetization axes and magnetic anisotropies depending on the type of RE in REBCO system, tri-axial magnetic crystal orientation under various MRF conditions for “twin-free” RE2Ba4Cu7Oy (y~15) and REBa2Cu4O8 superconductors are described as the second topic. The third topic is magnetic bi-axial crystal orientation in RE123. The behaviors of RE123 particles with twin microstructures under MRF are described, and one can understand that the bi-axial orientation using MRF can be achieved even for the existence of the twin microstructures. In the fourth topic, a combination of sheet-casting and uni-axial magnetic alignment under static field for another practical superconductor, (Bi,Pb)2223, is described.

PRINCIPLE OF MAGNETIC ALIGNMENT Uni-Axial Alignment of the First Easy Axis by SF Supposing that the alignment of particles under a static field of B obeys the Boltzmann distribution on the basis of Eq. (1), the probability of magnetically aligned particles, P(E), as a function of tilt angle () for the static field direction is expressed as follows:


Magnetic Alignment Techniques for HTSC 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 0 𝑘𝐵 𝑇

𝑃(𝐸𝜃 ) ∝ exp {2𝜇

− 𝜒ℎ

𝑟𝑑 )𝑐𝑜𝑠

2

𝜃}

157

⋯

Here, = easy - hard and easy>hard. Furthermore, for

(2) 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 2𝜇0 𝑘𝐵 𝑇

− 𝜒ℎ

𝑟𝑑 )

= Eq. (2) is rewritten as the following equation. 𝑃(𝐸𝜃 ) ∝ exp(−𝛼𝑠𝑖𝑛2 𝜃) ⋯ (3)  The relationship between P and  for  = 1, 10, 100 and 1000 is shown in Figure 2. When the degree of orientation of the first easy axis below 10 degrees is achieved, >1000 is required. Therefore, Eq. (1) is rewritten to the following equation for achieving the above orientation degree: 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 2𝜇0

− 𝜒ℎ

𝑟𝑑 )

> 1000𝑘𝐵 𝑇

(4)

Here, // and  in Eq. (1) are rewritten as easy and hard, respectively. For instance, in a sphere-shaped substance with 𝜒𝑒 𝑠𝑦 − 𝜒ℎ 𝑟𝑑 = 8.0 × 10−5, a diameter required for uniaxial alignment of the first easy axis under 1 T is approximately 0.6 micrometer at T = 300 K. This means that magnetic alignment is applicable even for feeble magnetic materials under strong magnetic fields.

Figure 2. Relationship between P(E) and  for various  values in the case of magnetic alignment under SF.



158

Uni-Axial Alignment of the Hard Axis by RF In the case of magnetic alignment using a rotating magnetic field (RF), the hard axis of magnetization is uniaxially aligned perpendicular to the rotating plane of the magnetic field. The energy of magnetic alignment for RF is expressed as follows: 2

E

　 VBrot (  easy -  hard ) 2 0

Brot 

　B 2

   (5)

(5)

   (6)

(6)

Here, Brot is a time-averaged magnetic field [20]. Therefore, the following equation is obtained from Eq. (5) and Eq. (6):

E

　 VB

2

4 0

(  easy -  hard )

   (7)

(7)

When Eq. (7) is compared to the left-hand side of Eq. (4), the orientation energy of RF is two times smaller than that of the static field. As explained in the previous section, the relationship between P(E) and  for the magnetic alignment by RF on the basis of the Boltzmann distribution is provided as follows: 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 0 𝑘𝐵 𝑇

𝑃(𝐸𝜙 ) ∝ exp {4𝜇

− 𝜒ℎ

𝑟𝑑 )𝑐𝑜𝑠

2

𝜙} ⋯

Here,  is the tilt angle for the direction normal to RF. For

(8) 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 4𝜇0 𝑘𝐵 𝑇

− 𝜒ℎ

𝑟𝑑 )

=

Eq. (8) is rewritten as follows: 𝑃(𝐸𝜙 ) ∝ exp(−𝛽𝑠𝑖𝑛2 𝜙) ⋯

(9)

 The relationship between P and  for  = 1, 10, 100 and 1000 is shown in Figure 3. If the degree of orientation of the hard axis below 10 degrees is achieved, >1000 is required. Therefore, the following equation is a requirement for achieving the above orientation degree condition under RF: 𝑉𝑩2 (𝜒𝑒 𝑠𝑦 4𝜇0

− 𝜒ℎ

𝑟𝑑 )

> 1000𝑘𝐵 𝑇


(10)

Magnetic Alignment Techniques for HTSC

159

Figure 3. Relationship between P(E) and  for various  values in the case of magnetic alignment under RF.

Tri-Axial Alignment by MRF The principle of the tri-axial grain orientation using MRF is explained in this section. Figure 4 shows a schematic of an intermittent type of MRF. Because the authors’ cryogen-free solenoidal superconducting magnet shows approximately 300 kg in weight, rotation control of the specimens is more realistic rather than that of the “heavy” magnet to generate MRFs. Therefore, a powder sample, which is RE-Ba-Cu-O powders mixed with epoxy resin, is horizontally rotated at two different steps in a static field (B) of B = 10 T applied along the transverse direction. At the angles of 0° and 180°, the powder sample is rested for t1 = 2 s, whereas the rotation process with  = 60 rpm is applied at the other angle regions. The original angle, 0°, of the powder sample is defined with regard to a direction normal to the  plane (see Figure 4) of the powder sample, which is parallel to the transverse B direction. The  plane is parallel to the powder sample rotation plane and the  plane is normal to both the  and  planes. If powders are triaxially oriented under MRF, the first easy, secondary easy, and hard axes of magnetization are aligned normal to the , , and  planes, respectively. As described in the previous section, the magnetic alignment at room temperature is achieved when the orientation energy exceeds the thermal energy at room temperature in principle. However, when the effect of thermal fluctuation which obeys the Boltzmann distribution is taken into consideration, a 1000 times larger orientation energy compared to the thermal energy (kBT) at room temperature is required for the accomplishment of the degrees of orientation within 10°. If the difference in the magnetic susceptiblities



160

between the first easy axis (1 axis) and the second easy axis (2 axis) and the difference in the magnetic susceptiblities of the second easy axis (2 axis) and the hard axis (3 axis) are expressed as 12 ( = 1 -2) and 23 ( = 2 -3), respectively, the magnetic alignment requirement can be described by Eqs. (11) and (12): 　 12V ( B1  B2 )  1000k BT 2 0

(11)

　  23V ( B2  B3 )  1000k BT 2 0

(12)

2

2

2

2

Note that respective B1, B2, and B3 represent the time-averaged magnetic fields perpendicular to the α, β, and γ planes in MRF. In the case of the intermittent type of MRF in Figure 4, these three values of B1, B2 and B3 are expressed as follows using the horizontal magnetic field B ( = 0H) and the resting time (t1) and rotation time (t2) in a period of the intermittent type of MRF: B1 

2  1   Bt1  Bt 2  2t1  t 2   

(13)

B2 

2 1   Bt 2  2t1  t 2   

(14)

B3  0

(15)

Equation (11) represents the requirement condition for magnetic separation of the first easy axis (1 axis) and the second easy axis (2 axis), and Eq. (12) gives the condition for separation of the second easy axis (2 axis) and the hard axis (3 axis).

Figure 4. A schematic figure of the relationship among B, measured surfaces of XRD (, , ), and the modulated rotation of a sample.



161

BEHAVIORS OF Y-BA-CU-O POWDERS UNDER MRF Crystal Structures and Twin Microstructures in Y-Ba-Cu-O The principle of tri-axial magnetic grain orientation using the MRF was described in the previous section. In this section, the magnetic alignment of Y-Ba-Cu-O compounds under the MRF are focused on as fundamental substances of RE-Ba-Cu-O. However, in order to understand the magnetic alignment of Y-Ba-Cu-O powders, information on crystal structures and twin microstructures in RE-Ba-Cu-O system is important. This is because the symmetry of crystal structures and the existence of twin microstructures directly lead to anisotropy of magnetic susceptibility (or magnetization) in molecular and grain levels. In addtion to the crystal structure of RE123 (see Figure 1), Figure 5 shows crystal structures of other two Y-Ba-Cu-O superconductors; YBa2Cu4O8 (Y124) and Y2Ba4Cu7Oy (Y247) with y = 15. All the three compounds are characterized by the layered perovskite structure with the possession of the Cu-O chain (parallel to the b-axis). The structural difference in these three compounds originates from the difference in the stacking sequence of the Cu-O chain along the c-axis; the repetition of a single-chain, that of the double-chain and the alternative repetition of a single- and double-chain for the Y123, Y124 and Y247 compounds, respectively. The double chain does not exhibit oxygen nonstoichiometry, whereas the single Cu-Oz chain shows large oxygen nonstoichiometry with 0 c at room temperature, which is in quite a contrast to the results in Y247 and Dy247. It is obvious that the conversion of the magnetization axes from c > a > b to b > a > c is due to the doping of Er. Note that RE247 (RE = Pr3+ [37] and Ho3+ [30] with J < 0) and Eu247 [30] containing Eu3+ with J < 0 at room temperature show c > a > b and b > a > c at room temperature, respectively. Figure 10 shows XRD patterns at the , , and planes for RE124 (RE = Dy and Er) powder samples aligned under the MRF with 10 T,  = 60 rpm and t1 = 2s in epoxy resin at room temperature as another example of twin-free RE-Ba-Cu-O powder samples. It is found from Figure 10 that the magnitude relationships of magnetization for Dy124 and Er124 are c > a > b and b > a > c at room temperature, respectively. This relationship is coincident with that for RE247 (RE = Dy and Er) in Figure 9. Other RE124 compounds (RE = Sm, Eu, Gd, Ho, Tm and Yb) can be grown by the flux method and the RE124 powder samples aligned under the MRF of 0Ha = 10 T in epoxy resin at room temperature are obtained as well [27]. From the XRD analysis for the RE124 powder samples, it is found that c > a > b is shown for RE = Sm and Ho and b > a > c is shown for RE = Eu, Gd, Tm, and Yb.


168


Figure 9. XRD patterns at ,  and  planes for the magnetically aligned power samples of (a)Dy247 and (b)Er247.

Figure 10, XRD patterns at ,  and  planes for the magnetically aligned power samples of (a)Dy124 and (b)Er124.

In order to understand grain-orientation effects of twinned powders under the MRF, XRD patterns at the , , and planes for RE123 (RE = Dy and Er) powder samples aligned under the MRF with 10 T,  = 60 rpm and t1 = 2s in epoxy resin at room temperature are shown in Figure 11. The enhancement of (00l) peaks appears at the and planes for the powder samples of Dy123 and Er123, respectively. Respective (h00) and (0k0) peaks are enhanced at the  and  planes for Dy123 and at the and  planes for Er123, and the (110) peak appeared at these planes simultaneously. It is found from the previous section that the appearance of both (h00) and (0k0) peaks means the bi-axial alignment of RE123 grains with twin microstructures. However, the appearance of the (110) peak suggests the existence of the uniaxial (c-axis) grain-orientation. The



169

magnitude relationships of magnetization for Dy123 and Er123 in the molecular level are c > a > b and b > a > c at room temperature, respectively. These relationships are coincident with those for RE124 and RE247 with RE = Dy and Er. Other RE123 (RE = Nd and Sm) powder samples aligned under the MRF of 10 T in epoxy resin at room temperature are obtained as well [38]. From the XRD analysis for the powder samples, it is found that c > a > b is shown for RE = Nd and Sm. That is, the bi-axial grainorientation of RE123 powders with twin microstructures can be achieved under the MRF of 10 T at room temperature.

Figure 11. XRD patterns at ,  and  planes for the magnetically aligned power samples of (a)Dy123 and (b)Er123.

Table 2. Relationship of magnetization axes for twinned RE123, twin-free RE247 and twin-free RE124. c > a > b and b > a > c are abbreviated as c >a >b and b >a >c, respectively. “UD” means an undetermined datum

J (RT) RE123 (twinned) RE247 (twin-free) RE124 (twin-free)

J(RT) RE123 (twinned) RE247 (twin-free) RE124 (twin-free)

Y 0 c >a >b c >a >b c >a >b Dy − c >a >b c >a >b c >a >b

Pr − UD c >a >b UD

Nd − c >a >b UD UD

Ho − c >a >b c >a >b c >a >b

Sm + c >a >b UD c >a >b Er + b >a >c b >a >c b >a >c

Eu + UD b >a >c b >a >c Tm + UD UD b >a >c


Gd 0 UD UD b >a >c Yb + UD UD b >a >c

170


Table 2 shows the relationships of magnetization axes for twinned RE123, twin-free RE124, and twin-free RE247 at room temperature together with the signs of the second order Stevens factors (J) for RE3+ ions [36] at room temperature. Although there is some undetermined data, it is found that the magnetization axes of RE-Ba-Cu-O in the molecular level depend only on the type of RE ions and are independent of the existence of twin microstructures. Furthermore, the first easy axes (or the hard axes) of magnetization are largely affected by the sign of J. One can qualitatively see c > a > b for RE3+ with J < 0 and b > a > c for RE3+ with J > 0. The magnetization axes in RE-Ba-Cu-O can be controlled by the choice of RE.

Tri-Axial Magnetic Anisotropies Depending on RE In the previous section, it was described that a dominant factor of the magnetization axes in RE-Ba-Cu-O was the type of RE. However, quantitative information on tri-axial magnetic anisotropies of RE-Ba-Cu-O in the molecular level is more important to evaluate the magnetic fields required for achieving high degrees of tri-axial (or bi-axial) orientation. Particularly in RE123, one should take the effects of the twin microstructures into account additionally for the appropriate choice of RE and the precise estimation of their tri-axial magnetic anisotropies in the grain level. In this section, semi-quantification of the tri-axial magnetic anisotropy of RE-Ba-Cu-O (or RE ion) in the molecular level is demonstrated. Here, we focus on (RE’1-xRE”x)Ba2Cu3Oy (y ~ 7) compounds [(RE’, RE”)123] containing RE’ and RE” ions with opposite signs of J at room temperature. From the conversion of magnetization axes by the replacement of RE,” one can experimentally determine two different critical RE”-doping levels, at which the magnitude relationships of paramagnetic susceptibilities between the c- and a-axes (xcc-a) and between the a- and b-axes (xca-b), are converted. Note that xcc-a is the RE”-doping level required for the conversion of the easy axis from the c-axis to the a-axis, and xca-b is the RE”-doping level required for the conversion of a relative easy axis from the a-axis to the b-axis. The ratio of magnetic anisotropy between RE’123 (RE’123) and RE”123 (RE’’123) is expressed using xc as follows:

RE’123/RE’’123 = xc / (1- xc)

(16)

The value of xc is intrinsic for each (RE’, RE”) 123 and is independent of the existence or absence of the twin microstructures. Furthermore, it should be emphasized that magnetic anisotropy ( = c - a or a - b) could be determined quantitatively from the RE” content (xc) required for the conversion of the magnetization axes without the determination of accurate mean radii of the source powders. As another approach for the



171

quantification of the  value, usage of relationship magnetic alignment energy and thermal energy at room temperature using a minimum magnetic field required for the achievement of magnetic grain orientation with the highest degrees of orientation is effective in principle. The magnetic alignment energy is proportional to  and the cube of the mean radius; therefore the accuracy of the mean radius is obviously sensitive to the accuracy of the obtained value of . On the other hand, in the case of the determination process of  in this chapter, the accuracy of a ratio of magnetic anisotropy is dominated only by the accuracy of the RE” content (xc) required for the conversion of the magnetization axes in (RE’, RE”)123. Therefore, precise information on the magnetic anisotropies of RE’123 and RE”123 in the molecular level can be obtained with this approach. Figures 12(a), 12(b), and 12(c) show XRD patterns at the , , and  planes for the (Y,Er)123 powder samples aligned under the MRF with 10 T,  = 60 rpm and t1 = 2s as an example to understand systematic changes of three different magnetization axes in (RE’, RE”)123 as a function of the doping level (x) of RE’. In the case of the results at the  plane in Figure 12 (a), the (00l) peaks are clearly enhanced for 0 ≤ x ≤ 0.06, and the (h00) and (0k0) peaks are enhanced for 0.05 ≤ x ≤ 1. This means that the first easy axes of (Y, Er)123 are converted from the c-axis to the b-axis by a slight doping of Er, and leads to xcc-a ~ 0.06. As shown in Figure 12(b), the second easy axes determined from XRD patterns at the  plane are insensitive to the doping level of Er. The (h00) and (0k0) peaks are enhanced independently with x and the magnitude relationship of intensities (I) between these two peaks always shows I(h00)>I(0k0). On the other hand, the (110) peak, which is a signal on the existence of the uni-axial (c-axis) aligned grains as explained in the previous section, is enhanced for 0.02 mab*) originated in the thick blocking layers in their crystal structure as shown in Figure 13. Due to such intrinsic characteristics, the anisotropic Jc ratio, Jc(//ab)/ Jc(//c) is quite large and it is considered to be larger than 100 for Bi2212. Therefore, formation of the c-axis aligned microstructure is necessary for achieving high Jc in polycrystalline materials. Silver sheathed tapes with multi-filaments of partially lead substituted Bi2223


174


((Bi,Pb)2223) have been well developed thus far and extensively used for superconducting cables, magnets, current leads and motors. The principal reason for the development of (Bi,Pb)2223 commercial tapes is the relatively easy formation of the caxis aligned (Bi,Pb)2223 filaments all through the long tape by flat rolling after formation of thin plate-like crystals of (Bi,Pb)2223 with a wide ab-plane in the first sintering process. During the flat rolling, cleavage of the crystal along the ab-plane easily occurs at very weakly bonded Bi-O double layer, which also contributes to form a dense and c-axis oriented microstructure. The Jc of superconducting filaments of commercial (Bi,Pb)2223 tapes is approximately 5 x 104 A cm-2 at 77 K in low fields, which is two orders of magnitude higher than that of randomly grain oriented sintered bulks, however, much lower than that of eptaxially grown thin films [39] and the local part of the tape [40]. This means that the ideally dense and grain oriented microstructure is partially formed in the filaments and the degree of grain orientation is not enough for high Jc as in the thin films. Therefore, room remains to improve critical cuttent performance of (Bi,Pb)2223 tapes.

Figure 13. Crystal structure of (Bi,Pb)2223.



175

uniaxial press  sintering

magnetic alignment  uniaxial press  sintering

Figure 14. Surface XRD patterns of (Bi,Pb)2223 bulks sintered at 815°C for 8 h in PO2 = 10 kPa and a total gas pressure of 10 MPa. There patternsFig. were taken after removal of top surface region with a 14 Horii thickness of ~20 m.

(a) uniaxial press  sintering

(b) magnetic alignment  uniaxial press  sintering

Fig. 15 Horii Figure 15. Secondary electron images of fractured cross sections of (Bi,Pb)2223 bulks sintered at 815°C for 8 h after uniaxial pressing, (a) without magnetic orientation process, (b) with magnetic alignment process by slip-casting under 10 T.


176


The magnetic easy axis direction of (Bi,Pb)2223 is parallel to the the c-axis, because magnetic anisotropy at the CuO2 plane is predominant in the crystal, and hence, the c-axis direction of (Bi,Pb)2223 grain is controlled by applying a static magnetic field. In this case, the c-axis of (Bi,Pb)2223 tends to be parallel to the direction of the external field. Combining these features of this compound, plate-like crystal shape with wide ab-plane, easy cleavage along the ab-plane and c-axis orientation by applying static fields, magnetic grain alignment for (Bi,Pb)2223 powder and uni-axial pressing and/or flatrolling are a promising way to obtain dense and c-axis aligned green pellets of (Bi,Pb)2223. Figure 14 shows surface XRD patterns for Bi(Pb)2223 sintered bulks after removal of the top surface region with ~20 m in thickness by polishing. Since the c-axis grain orientation occurs at the top surface region up to a depth of ~20 m by uniaxial pressing to prepare pellets, strong reflections of (00l) are observed in both patterns. By comparing these patterns carefully, diffraction peaks except for (00l) are suppressed in a sample magnetically aligned before uniaxial pressing. The magnetically grain alignment is performed under 10 T applied vertical direction to the disk-shaped stage of slip-casting using a slurry of (Bi,Pb)2223 powder and etahnol at room temperature. Sintering after uniaxial pressing was carried out by using a hot-isostatic pressing furnace under a total gas pressure of 10 MPa containing a partial oxygen pressure of 10 kPa at 815°C for 8 h. Samples were sealed in silver foil before sintering. Details of the experiment can be found elsewhere [41]. Secondary electron images of fractured cross sections of these samples are shown in Figure 15. A dense and strongly c-axis oriented microstructure is observed in the bulk prepared after magnetic alignment, whereas grain alignment is apparently poor in the bulk without the magnetic alignment process. These results mean that the c-axis grain aligned texture is maintained even after high temperature sintering. Although grain coupling of (Bi,Pb)2223 in the dense and strongly c-axis oriented sample is not sufficiently strong for carrying large superconducting current due to poor chemical coupling between preliminary synthesized (Bi,Pb)2223 crytals, this method can be usuful for the development of (Bi,Pb)2223 bulks and tapes with high Jc when we start from a powder mixture of (Bi,Pb)2223 and its precursor materials.

SUMMARY The magnetic alignment technique for the tri-axial (or bi-axial) alignment of HTSC was focused on as the first topic in this chapter. The tri-axial magnetic alignment, which is achieved by MRFs, and exhibits advantageous points of room temperature and nonvacuum processes. It is quite different from the fabrication process of coated conductors using an epitaxial growth technique, and we introduced behavior of practical HTSC; RE-



177

Ba-Cu-O powders with and without the twin microstructures under the MRF and (Bi,Pb)2223 under SF. In the case of RE-Ba-Cu-O, powders of all three systems, RE124 and RE247 without twin microstructures and RE123 with the twin microstructures, could be tri-axially (or biaxially) aligned under the MRF of 10 T in epoxy resin at room temperature. On the other hand, the existence of the twin microstructures in RE123 reduced the inplane orientation degrees, and the orientation degrees and the magnetization axes depended on the types of RE. Heavy RE ions, such as Dy, Ho and Er, showed higher magnetic anisotropies at room temperature. The findings in this chapter indicate that the tri-axial magnetic alignment technique enables quantification of three-dimensional magnetic anisotropy even from powder samples, and that the enhancement of magnetic anisotropy through the appropriate choice of the RE’(RE”) ion in RE-based cuprate superconductors leads to the reduction of the magnetic field required for the production of tri-axially oriented bulks and thick films based on combinations of the magnetic alignment and colloidal techniques [42-46]. Therefore, magnetic alignment using MRF has the potential to be a low-cost production technique of RE123 bulk magnets and superconducting cables without using epitaxy technology, precise control of growth temperature, and a highvacuum system. As another topic, we introduced the magnetic alignment process in the fabrication of (Bi,Pb)2223 bulks at room temperature and highly dense and strongly uni-axially (c-axis) aligned bulks were obtained.

ACKNOWLEDGMENTS The authors thank Prof. Toshiya Doi, Mr. Shotaro Fujioka, Mr. Tomohiro Nishioka, Mr. Itsuki Arimoto, and Mr. Daisuke Notsu for their kind support in this work. This work was partly supported by Adaptable and Seamless Technology Transfer Program through Target-driven R&D (A-STEP) from Japan Science and Technology Agency (JST) and JSPS KAKENHI Grant Number 17H03235 from the Japan Society for the Promotion of Science (JSPS).

REFERENCES [1] [2] [3]

M. V. Berry and A. K. Geim, Euro. J. Phys. 18 (1997) 307. A. Geim, Physics Today 51 (1998) 36. M. K. Wu, J. R. Ashburn, C. J. Torng, P. H. Hor, R. L. Meng, L. Gao, Z. J. Huang, Y. Q. Wang, and C. W. Chu, Phys. Rev. Lett. 58 (1987) 908.


178 [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

[26]

Shigeru Horii and Jun-ichi Shimoyama H. Maeda, Y. Tanaka, M. Fukutomi, and T. Asano, Jpn. J. Appl. Phys. 27 (1988) L209. K. Kishio, J. Shimoyama, T. Kimura, Y. Kotaka, K. Kitazawa, K. Yamafuji, Q. Li, and M. Suenaga, Physica C 235-240 (1994) 2775. D. A. Cardwell, Mater. Sci. Eng. B 53 (1998) 1. S. R. Foltyn, L. Civale, J. L. MacManus-Driscoll, Q. X. Jia, B. Maiorov, H. Wang, and M. Maley, Nat. Mater. 6 (2007) 631. P. J. Hirschfeld, P. Wölfle, J. A. Sauls, D. Einzel, and W. O. Putikka, Phys. Rev. B 40 (1989) 6695. Y. Iye, T. Tamegai, H. Takeya, and H. Takei, Jpn. J. Appl. Phys. 26 (1987) L1057. D. Dimos, P. Chaudhari, and J. Manhart, Phys. Rev. B 41 (1990) 4038. D. Dimos, P. Chaudhari, J. Mannhart, and F. K. LeGoues, Phys. Rev. Lett. 61 (1988) 219. R. B. Laibowitz, R. H. Koch, P. Chaudhari, and R. J. Gambino, Phys. Rev. B 35 (1987) 8821. K. Sawano, M. Morita, M. Tanaka, T. Sasaki, K. Kimura, S. Takebayashi, M. Kimura, and K. Miyamoto, Jpn. J. Appl. Phys. 30 (1991) 1153. K. Sato, S. Kobayashi, and T. Nakashima, Jpn. J. Appl. Phys. 51 (2012) 010006. K. Watanabe, S. Awaji, T. Fukase, Y. Yamada, J. Sakuraba, F. Hata, C. K. Chong, T. Hasebe, and M. Ishihara, Cryogenics 34 (1994) 639. K. Watanabe and S. Awaji, J. Low Temp. Phys. 133 (2003) 17. Y. Ikezoe, N. Hirota, J. Nakagawa, and K. Kitazawa, Nature 393 (1998) 749. T. Kimura, F. Kimura, and M. Yoshino, Langmuir 22 (2006) 3464. M. Stains, J. Y. Genoud, A. Mawdsley, and V. Manojlovic, IEEE Tran. Appl. Supercond. 9 (1999) 2284. T. Kimura, Polym. J. 35 (2003) 823. J. D. Jorgensen, B. W. Veal, W. K. Kwok, G. W. Crabtree, A. Umezawa, L. J. Nowicki, and A. P. Paulikas, Phys. Rev. B 36 (1987) 5731. K. Kishio, J. Shimoyama, T. Hasegawa, K. Kitazawa, and K. Fueki, Jpn. J. Appl. Phys. 26 (1987) L1228. Z. Hiroi, M. Takano, Y. Takeda, R. Kanno and Y. Bando, Jpn. J. Appl. Phys. 27 (1988) L580. J. Karpinski, E. Kaldis, E. Jilek, S. Rusiecki, and B. Bucher, Nature 336 (1988) 660. S. Horii, Y. Yamada, H. Ikuta, N. Yamada, Y. Kodama, S. Katano, Y. Funahashi, S. Morii, A. Matsushita, T. Matsumoto, I. Hirabayashi, and U. Mizutani, Physica C 302 (1998) 10. Y. T. Song, J. B. Peng, X. Wang, G. L. Sun, and C. T. Lin, J. Cryst. Growth 300 (2007) 263.



179

[27] M. Yamaki, S. Horii, M. Haruta, and J. Shimoyama, Jpn. J. Appl. Phys. 51, 010107 (2012). [28] T. Fukushima, S. Horii, H. Ogino, T. Uchikoshi, T. S. Suzuki, Y. Sakka, J. Shimoyama, and K. Kishio, Appl. Phys. Express 1 (2008) 111701. [29] T. Fukushima, S. Horii, T. Uchikoshi, H. Ogino, A. Ishihara, T. S. Suzuki, Y. Sakka, J. Shimoyama, and K. Kishio, IEEE Trans. Appl. Supercond. 19, 2961 (2009). [30] S. Horii, S. Okuhira, M. Yamaki, K. Kishio, J. Shimoyama and T. Doi, J. Appl. Phys. 115 (2014) 113908. [31] S. Horii, M. Yamaki, H. Ogino, T. Maeda, and J. Shimoyama, Physica C 470 (2010) 1056. [32] Y. Iijima, N. Kaneko, S. Hanyu, Y. Sutoh, K. Kakimoto, S. Ajimura, T. Saitoh, Physica C 445-448 (2006) 509. [33] J. M. Ferreira, M. B. Maple, H. Zhou, R. R. Hake, B. W. Lee, C. L. Seaman, M. V. Kuric, and R. P. Guertin, Appl. Phys. A47 (1988) 105. [34] A. Ishihara, S. Horii, T. Uchikoshi, T. S. Suzuki, Y. Sakka, H. Ogino, J. Shimoyama, and K. Kishio, Appl. Phys. Express 1 (2008) 031701. [35] P. de Rango, M. Lees, P. Lejay, A. Sulpice, R. Tournier, M. Ingold, P. Germi, and M. Pernet, Nature 349 (1991) 770. [36] K. W. H. Stevens, Proc. Phys. Soc. London, Ser. A, 29 (1952) 209. [37] S. Horii, unpublished. [38] S. Horii, T. Nishioka, I. Arimoto, S. Fujioka, and T. Doi, Supercond. Sci. Technol. 29 (2016) 125007. [39] Y. Hakuraku and Z. Mori, J. Appl. Phys 73 (1993) 306. [40] A. Polyanskii et al. IEEE Trans. Appl. Supercond. 11 (2001) 3269-3270. [41] K. Obata, J. Shimoyama, A. Yamamoto, H. Ogino, K. Kishio, S. Kobayashi and K. Hayashi, Phys. Procedia 36 (2012) 665. [42] S. Horii, I. Matsubara, M. Sano, K. Fujie, M. Suzuki, R. Funahashi, M. Shikano, W. Shin, N. Murayama, J. Shimoyama, and K. Kishio, Jpn. J. Appl. Phys. 42 (2003) 7018. [43] Y. Zhou, I. Matsubara, S. Horii, R. Funahashi, M. Shikano, J. Shimoyama, K. Kishio, W. Shin, N. Izu, and N. Murayama, J. Appl. Phys. 93 (2003) 2653. [44] T. Okamoto, S. Horii, T. Uchikoshi, T. S. Suzuki, Y. Sakka, R. Funahashi, N. Ando, J. Shimoyama, and K. Kishio, Appl. Phys. Lett. 89 (2006) 081912. [45] T. S. Suzuki, T. Uchikoshi, and Y. Sakka, Sci. Tech. Adv. Mater. 9 (2006) 356. [46] S. Tanaka, A. Makiya, Z. Kato, N. Uchida, T. Kimura, and K. Uematsu, J. Mater. Res. 21 (2006) 703.





Chapter 7

MAGNETIZATION OF POLYCRYSTALLINE HIGH-TC SUPERCONDUCTORS Denis Gokhfeld Kirensky Institute of Physics, Federal Research Center KSC SB RAS, Krasnoyarsk, Russia

ABSTRACT Magnetization loops of polycrystalline high-Tc superconductors always are semireversible; they have some asymmetry relative to field axis. These magnetization loops are described successfully by the extended critical state model, which accounts the equilibrium magnetization of the grain surface. The model is applied to determine the intragranular critical current density, the depth of the equilibrium surface region and the grain size from magnetic measurements. The dependence of the critical current density on the magnetic field, the pinning force scaling, the full penetration field and the irreversibility field are discussed.

Keywords: magnetization, critical current, pinning force, surface, scaling

1. INTRODUCTION External magnetic field induces circulating supercurrents in a superconductor. At the same time the circulating supercurrent hinders the magnetic field to fill the whole superconductor. This is reason of the superconducting diamagnetism. Magnetic flux can enter into type-II superconductor as Abrikosov vortices [1]. Pinning of the Abrikosov vortices on different defects produces a magnetization hysteresis. The Bean’s critical


182

Denis Gokhfeld

state model [2] is used during long time to analyze the magnetization M of type-II superconductors. In the critical state model the density of circulating supercurrent is equal to the critical current density jc. The value of jc determines the magnetic flux gradient and the depth of magnetic flux penetration. The critical state model qualitatively describes the magnetization hysteretic dependencies of hard superconductors (superconductors with strong pinning). Quantitative agreement between computed curves and experimental magnetization loops is achieved with using a field dependent jc [3, 4]. Magnetization loops of the hard superconductors are nearly symmetric along magnetic field axis H. Polycrystalline high-Tc superconductors are not hard. Their magnetization loops demonstrate the crossover from a nearly symmetric hysteresis to a reversible curve with decreasing temperature. Such the asymmetric magnetization loops are adequately described in the framework of the extended critical state model (ECSM) [5, 6, 7]. In ECSM, the total magnetization of a sample is emerged from the surface magnetization and the nonequilibrium magnetization of the remaining volume of the sample. The asymmetry of the magnetization loop is determined by the share of the equilibrium surface magnetization in the total magnetization of the sample. ECSM is used to describe and analyze magnetization loops of polycrystalline superconductors. The model already was applied to study various polycrystalline compounds: Re-123 (Re = Y, Nd, or Eu) [8, 9, 10, 11], Bi-2223 [6], MgB2 [12], Ba0.6K0.4BiO3 [13], high-porous Bi-2223 [6] and Bi-2212 nanowire fabrics [14, 15]. This chapter is organized as follows: (i) the extended critical state model and its application for polycrystalline samples are described (Section 2), (ii) the field dependencies of the critical current and the pinning force are discussed, and (iii) meanings and typical values of used parameters are listed (Section 3).

2. MODEL 2.1. Magnetic Flux Profiles and Magnetization The extended critical state model (ECSM) is a modification of the critical state model. ECSM was applied by Chen and coauthors [5] to describe asymmetric magnetization loops of high-Tc superconductors. In ECSM, a specific equilibrium magnetization of the sample surface is accounted. Abrikosov vortices are not pinned in the surface region with the depth ls, which is about the London penetration depth λ. The equilibrium magnetization of the surface is the reason for the observed asymmetry of magnetization loops. The magnetization of the inner bulk is described by the critical state model [2] with modified boundary conditions taking into account the magnetization of the surface layer. Variations of ls with the magnetic field H was not accounted in Chen’s


Magnetization of Polycrystalline High-Tc Superconductors

183

article [5]. In presented work we consider how the magnetization loops are affected by the ls(H) dependence. The depth ls is determined by an interaction magnetic flux in the sample with the sample surface [16, 17]. The phenomenological ls(H) dependence is suggested here: 𝛽

𝑙𝑠 (𝐻, 𝑇) = 𝑅 − (𝑅 − 𝑙𝑠0 (𝑇)) (1 −

|𝐻| ) , 𝐻irr (𝑇)

(1)

where ls0 is the value of ls at H = 0, Hirr is the irreversibility field, and β is a positive constant, β ≤ 1. Equation (1) qualitatively agrees with the λ(H) dependency [18]. Further in this section we follow to the method described in works [19, 20] to find the field distribution in a superconducting sample. The magnetization M(H) is defined as M(H) = – H + 𝐵̅(H)/μ0, where 𝐵̅ is the averaged magnetic field inside the sample and μ0 is the magnetic constant. We consider the cylindrical sample with the length much longer than the radius R. The external magnetic field is applied along the longer cylinder axis Thus we neglect demagnetization effects and field inhomogeneity along the longer axis. This task is one-dimensional: all physical quantities depend on the distance r from the cylinder axis. The considered approach can be generalized for samples with other forms, e.g., for the sample with the rectangular cross-section [7, 19]. The magnetization of the cylinder is determined by the following expression:

𝑀(𝐻) = −𝐻 +

𝑅 2 ∫ 𝑟𝐵𝑑𝑟, 𝜇0 𝑅 2 0

(2)

The distribution of the flux density B inside the superconducting sample is determined by 𝑑𝐵 𝑑𝑟 = ± 𝜇0 𝑗𝑐 (𝐵),

(3)

where jc is the local critical current density in the sample. The surface supercurrent density js is determined by the surface barrier. Focusing on the behavior of the magnetization loops in strong magnetic fields, we assume js = jc. Then the boundary condition of Eq. (3) is B = μ0H at r = R. The solution of Eq. (3) can be obtained in the form: ±𝐵 = 𝐹 −1 (𝐹(𝐵0 ) ± 𝜇0 𝑗𝑐0 (𝑅 − 𝑟 − 𝑟0 )),

(4)

𝐵

where the function F(B) is defined as F(B) = ∫0 𝑗𝑐0 𝑗𝑐 (𝐵′)d𝐵′ with the inverse function F-1, and jc0 is the value of jc(B) at B = 0. The integration constant is chosen to have B = B0 at r = R–r0. The required jc(B) dependence is considered in Section 2.2.


Denis Gokhfeld

184

Hp

B D

H

D

A

A E

B

B

C

0

C

G F

F

0

R R-ls

R-ls R

r Figure 1. Magnetic flux distribution in the cylindrical sample.

The magnetic flux firstly comes into the surface layer (plot AB in Figure 1). Plot BC corresponds to penetration of the Abrikosov vortices into the inner region of the sample. When the external field H equals the full penetration field Hp, the magnetic flux reaches the center of the sample. With a further increase in the magnetic field H, the flux density B at the center of the sample increases (point E). When the external field decreases after the maximum reached value Hm, the trapped flux remains in the central region of the sample (plot BDE). This trapped magnetic flux leads to the observed hysteresis of the M(H) dependence. Plot CFG in Figure 1 corresponds to the trapped flux in the sample after the external field H reaching –Hm. Table 1 contains the boundary conditions for the B(H,r) profiles. Bs(H) is the value of B at the depth ls(H) from the surface, it is given by Bs(H) = F-1(F(μ0H) – μ0 jc0 ls(H)). Table 1. Boundary conditions for B(H,r) profiles

AB BC BD DE CF FG

B0

r0

μ0H Bs(H) Bs(H) Bs(Hm) Bs(H) Bs(Hm)

0 ls(H) ls(H) ls(H) ls(H) ls(H)

Sign before μ0 jc (R–r–r0) – – + – – –


Sign of B + + + + – –


185

3

R >> ls0

Bc/0Hp

R = 2ls0

0

-3 -3

0

3

H/Hp Figure 2. Flux density Bc in the middle of the sample. The axes are normalized to the full penetration field Hp.

Figure 1 demonstrates that the flux density profile in the central region depends on the magnetization prehistory. Accordingly, there are three branches of the M(H) dependence: (i) the branch of the initial magnetization with the increase of H from zero to Hm corresponds to the profile ABC in Figure 1; (ii) the branch M+(H) with the decrease of H from Hm to 0 corresponds to profile ABDE in Figure 1 (the symmetric branch is resulted with H changed from –Hm to 0); and (iii) the branch M–(H) with H increased from 0 to Hm after the circulation of the external field from 0 to –Hm and back to 0 corresponds to the profile ABCFG in Figure 1 (the symmetric branch is resulted with H changed from 0 to –Hm). The full penetration field Hp and the flux density in the middle of the sample Bc are calculated from Eq. (4). Hp is determined as Hp = F-1(μ0 jc0R)/μ0. An approximated formula for Hp is given in Section 2.2. The upper limit of Hp is Hp = jc0R, that approaches for small R. The central flux density Bc is related to the trapped field in the sample. The Bc(H) dependence is hysteretic. It is determined by Eq. (4) with r = 0. The Bc(H) loops are plotted in Figure 2 for the gross sample (R ⪢ ls0) and for the small one (R = 2ls0). The maximal value of Bc at H = 0 occurring for R ⪢ ls is Bc = μ0Hp. The trapped flux Bc(0) decreases with increasing the ratio ls/R.

2.2. jc(B) Dependence To fit experimental magnetization loops, one should select an appropriate dependence of the critical current density jc on B. As B grows from 0 to the upper critical


Denis Gokhfeld

186

field Bc2 = μ0Hc2, the jc(B) dependence decreases from jc0 to 0. The peak effect can distort the decreasing jc(B) dependence that is considered in [11, 13, 21, 22]. Various dependences jc(B) were tested in the calculations of the magnetization loops. The M(H) dependence calculated using the Bean model [2] (jc(B,T) = jc0(T)) does not reproduce well experimental magnetization loops. The Kim relation jc(B,T) = jc0(T)/(1+|B|/B0(T)) [3], which provides decreasing of jc with local field, gives a good fit of M(H) dependencies in small fields H 1. However, as we see hereinafter (Section 3.2), the DewHughes scaling law [24] requires α < 1. The approaching of the magnetization to 0 is well described with a combination of the Anderson-Kim and the exponential formulas [7]: 𝑗𝑐 (𝐵, 𝑇) =

𝑗𝑐0 (𝑇)

, (|𝐵| 𝐵0 (𝑇)) + exp(|𝐵| 𝐵1 (𝑇))

(6)

α

Urban [25] modified the Kim relation to describe both low field and high field regions of jc(B) dependence: 𝑗𝑐 (𝐵, 𝑇) = 𝑗𝑐0 (𝑇)

1 − |𝐵| 𝐵𝑐2 (𝑇) , 1 + |𝐵| 𝐵0 (𝑇)

(7)

A general form of all the listed dependencies is suggested here: 𝛼

𝑗𝑐 (𝐵, 𝑇) = 𝑗𝑐0 (𝑇)

1 − (|𝐵| 𝐵𝑐2 (𝑇))

𝛼

1 + (|𝐵| 𝐵0 (𝑇))

(8) ,

This jc(B) dependence accounts the Urban deflection and satisfies the Dew-Hughes scaling law. The function (8) is the general form of the Bean dependence (with α = 0), the Kim dependence (with α = 1), and the exponential dependence (with B/B0 Hirr. In equation (9) H0=B0/μ0 and γ is the positive coefficient such that γ = nβ. The function (9) tends to 0 as H approaches to Hirr(T) so it decreases faster than the jc(μ0H) dependence (8). Formula (9) describes well any experimental Jc(H) dependencies without the peak effect. Figure 4 demonstrates the Jc(H) curves, calculated by Eq. (9) with the same parameters as the magnetization loops in Figure 3. The Jc(H) dependencies determined from the M(H) loops using the Bean model are shown in Figure 4 also (points). The Bean model [2] establishes relation between the irreversible magnetization ΔM(H) = M+(H) – M–(H) and the Jc(H) dependence: Jc(H) = ΔM(H)/(kR), where the coefficient k depends on the sample geometry, k = 2/3 for the cylindrical sample. It is seen in Figure 4, that the Bean formula underestimates Jc(H) in the low field region. For H ≥ Hp, the curves of ECSM model gives surprisingly good agreement with the Jc(H) dependencies determined from the M(H) loops using the Bean model. As it was tested for various computed magnetization loops, the good coincidence keeps for any values of R. The reason for this quantitative coincidence is the same influence of the surface on the macroscopic critical current and on the irreversible magnetization ΔM(H). The Jc(H) dependence of a smaller sample decreases faster than the Jc(H) dependence of a larger sample. This is due to omitting the surface region with the field dependent depth ls(H). The irreversible


Denis Gokhfeld

190

magnetization of smaller samples also decreases faster than ΔM(H) of large samples. As the numerical calculations display, the curves Jc(H) and ΔM(H) decrease identically and both equal 0 at H ≥ Hirr. The dependence of the pinning force density on the applied magnetic field is determined by Fp(H) = μ0H Jc(H). Given H ⪢ H0 and H ⪡ Hc2, one obtains the pinning force density from Eq. (9): 𝛾

𝐹𝑝 (𝐻, 𝑇) ≈ µ0 𝐽𝑐0 (𝑇)𝐻0 (𝑇)

|𝐻|1−𝛼 |𝐻| − (1 ) , 𝐻0 (𝑇)1−𝛼 𝐻irr (𝑇)

(10)

This result coincides with the well-known Dew-Hughes scaling law [17]: 𝐹𝑝 (𝐻, 𝑇) ℎ𝑝 (1 − ℎ)𝑞 = 𝑝 , 𝐹𝑝0 (𝑇) ℎ0 (1 − ℎ0 )𝑞

(11)

where h0 is the position of the maximum, h0 = p/(p + q), p = 1 − α and q = γ. The formula (10) can be obtained from both the jc(B) dependencies (5) and (8). For polycrystalline superconductors the positive coefficient p is more often smaller than 1 [27], so α is usually to be smaller than 1 too. With this condition the jc(B) dependence (8) fits better than (5).

3.3. Scale of Screening Current The size R, which gives the circulation region of the screening current, is related to the sample structure. For granular superconductors, there is an ambiguity in choosing the value of R [12, 28] because the supercurrents circulate on at different length scales. The actual value of R can change with H, so R may be the radius of the sample or the effective radius of the grains or the radius of clusters consisting of several grains. Both the width ΔM and the asymmetry of magnetization loop depend on R. The asymmetry of magnetization loop with respect to the H axis is determined by the ls0/R ratio. Noticeable asymmetry of hysteresis is observed for ls/R > 0.1. This gives R < 10λ that can be used for fast and rough estimations of the circulation scale in the polycrystalline superconductors. The formula for accurate estimation of R is suggested hereinafter. Let us write equation: γ

Δ𝑀(𝐻) 𝑙𝑠 (𝐻) = 𝑗𝑐 (µ0 𝐻) (1 − ) , 𝑘𝑅 𝑅


(12)


191

where the left side is the Bean model formula for the critical current density Jc(H), the right side is the short form of Eq. (9). The magnetization loops of the samples with R ⪢ ls are near symmetric relative to H axis in the wide field range (see curve for R = 7.5 μm on Figure 3). For such the loops ΔM(H) = 2|M–(H)| and 2|M–(H)|/kR = jc(μ0H). From here, neglecting variations of ls(H) at small H, we obtain estimation for the ls0/R ratio in polycrystalline superconductors: 1

Δ𝑀(𝐻𝑝 ) 3 𝑙𝑠0 ≈1−( ) . 𝑅 2|𝑀− (𝐻𝑝 )|

(13)

It is convenient to determine the ls0/R ratio at the point H = Hp, where the branch of virgin magnetization and the M–(H) branch join together. The sample porosity does not disturb the estimation (13). But this estimation is sensitive to a tilt of the magnetization loops. Therefore, before estimating, the magnetization MN(H) of additional magnetic phases should be subtracted from the M(H) loop. For the analyzed polycrystalline superconductors [6, 8-12], the magnetization loops are successfully described by using ECSM with R equaling the effective grain radius (at H > 100 Oe). For the Bi-2212 nanowire fabrics [14, 15], R was found to equal the averaged nanowire radius.

3.4. Parameters Parameters required to compute a M(H) loop are listed here. Suggested parameters are typical for polycrystalline Y-123, Bi-2212, and Bi-2223 superconductors. The value of the local critical current density jc can approach to the depairing current density of superconductors. For high-Tc superconductors jc0 is on the order of 1012-1013 A/m2 at T = 4.2 K. The parameter B0 determines the decreasing of jc(B) dependence. As it is supposed, B0 relates to the upper critical field Hc2 and B0(T) ~ 0.01 Bc2(T). For high-Tc superconductors B0 ~ 1 Tesla at T = 4.2 K. The parameter α is expressed from the Dew-Huges scaling law as α = 1 − γ/(1/h0 − 1). The value of h0 is easily determined from Fp(H) graphics and this gives α between 0 and 1. If h0 = 0.2, then α ≈ 0.25. The position h0 is influenced by anisotropy and percolation [29] as well as α. The high anisotropy of grains and the low number of percolation paths results in the decreased value of h0 [14, 29] and the increased value of α. The value of ls0 is about the London penetration depth λ [5, 16, 17]. It is about 100500 nm at T = 4.2 K.


192

Denis Gokhfeld

The irreversibility field Hirr relates to the averaged pinning potential and it depends on the type of the defects and the number of these pinning centers in the superconductor. The value of Hirr decreases with increasing the ratio ls0(T)/R, so Hirr ~ (R/ls0 – 1) for Hirr ⪡ Hc2. It should be noted that the parameters have different values for some special cases. So the value of ls0 was found to be a few times smaller than λ in Bi-2212 nanowires, where the nanowire radius is about λ [15].

CONCLUSION The extended critical state model explains the observed peculiarities of the magnetization loops of polycrystalline superconductors. The asymmetry of the magnetization hysteresis relative to the H axis relates to the equilibrium magnetization of the surface. Equation (4) of the model gives the magnetic flux distributions, the full penetration field and the central trapped field. The general form of jc(B) dependence is suggested (Eq. (8)). ECSM describes the connection of the macroscopic critical current density Jc(H) with the ls0/R ratio (Eq. (9)). The method is found to estimate the ls0/R ratio from the observed asymmetry of the magnetization loop (Eq. (13)).

ACKNOWLEDGMENTS The author is thankful to M. R. Koblischka (Saarland University, Saarbrücken, Germany), D. A. Balaev and V. V. Valkov (Kirensky Institute of Physics, Krasnoyarsk, Russia) for fruitful discussions.

REFERENCES [1] [2] [3] [4]

A. A. Abrikosov, On the magnetic properties of superconductors of the second group. Sov. Phys. JETP 5, 1174-1182 (1957). C. P. Bean, Magnetisation of hard superconductors. Phys. Rev. Lett. 8, 250-253 (1962). Y. B. Kim, C. F. Hempstead, and A. Strand, Magnetization and critical supercurrents. Phys. Rev. Lett. 129, 528-535 (1963). F. Irie and K. Yamafuji, Theory of flux motion in non-ideal type-II superconductors. J. Phys. Soc. Jpn. 23, 255-268 (1967).



[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16] [17]

193

D. X. Chen, R. W. Cross, and A. Sanchez, Effects of critical current density, equilibrium magnetisation and surface barrier on magnetisation of high temperature superconductors. Cryogenics 33, 695-703 (1993). D. M. Gokhfeld, D. A. Balaev, M. I. Petrov, S. I. Popkov, K. A. Shaykhutdinov, and V. V. Val'kov, Magnetization asymmetry of type-II superconductors in high magnetic fields. J. Appl. Phys. 109, 033904 (2011). D. M. Gokhfeld, An extended critical state model: Asymmetric magnetisation loops and sield dependence of the critical current of superconductors. Phys. Solid State 56, 2380-2386 (2014). Z. D. Yakinci, D. M. Gokhfeld, E. Altin, F. Kurt, S. Altin, S. Demirel, M. A. Aksan, and M. E. Yakinci, Jc enhancement and flux pinning of Se substituted YBCO compound. J. Mater. Sci.: Mater. Electron. 24, 4790-4797 (2013). E. Altin, D. M. Gokhfeld, F. Kurt, and M. E. Yakinci, Physical, electrical, transport and magnetic properties of Nd(Ba,Nd)2.1Cu3O7-d system. J. Mater. Sci.: Mater. Electron. 24, 5075-5084 (2013). E. Altin, D. M. Gokhfeld, S. Demirel, E. Oz, F. Kurt, S. Altin, and M. E. Yakinci, Vortex pinning and magnetic peak effect in Eu(Eu,Ba)2.125Cu3Ox. J. Mater. Sci.: Mater. Electron. 25, 1466-1473 (2014). D. M. Gokhfeld, D. A. Balaev, I. S. Yakimov, M. I. Petrov, S. V. Semenov, Tuning the peak effect in the Y1−xNdxBa2Cu3O7−δ compound. Ceram. Int. 43, 9985-9991 (2017). E. Altin, D. M. Gokhfeld, S. V. Komogortsev, S. Altin, and M. E. Yakinci, Hysteresis loops of MgB2 + Co composite tapes. J. Mater. Sci.: Mater. Electron. 24, 1341-1347 (2013). D. A. Balaev, D. M. Gokhfeld, S. I. Popkov, K. A. Shaikhutdinov, L. A. Klinkova, L. N. Zherikhina, A. M. Tsvokhrebov, Increase in the magnetization loop width in the Ba0.6K0.4BiO3 superconductor: possible manifestation of phase separation. J. Exp. Theor. Phys. 118, 104–110 (2014). M. R. Koblischka, D. Gokhfeld, X. L. Zeng, U. Hartmann, C. Chang, and T. Hauet, Pinning force scaling of electrospun Bi-2212 nanowire networks. Solid State Commun. 264, 16-18 (2017). X. L. Zeng, T. Karwoth, M. R. Koblischka, U. Hartmann, D. Gokhfeld, C. Chang, and T. Hauet, Analysis of magnetization loops of electrospun non-woven superconducting fabrics. Phys. Rev. Mater. 1, 044802 (2017). J. R. Clem, A model for flux pinning in superconductors. Low Temperature Physics-LT 13 (1974). L. Burlachkov, Magnetic relaxation over the Bean-Livingstone surface barrier. Phys. Rev. B 47, 8056-8064 (1993).


194

Denis Gokhfeld

[18] A. Maeda, Y. Iino, T. Hanaguri, N. Motohira, K. Kishio, and T. Fukase, Magneticfield dependence of the London penetration depth of Bi2Sr2CaCu2Oy. Phys. Rev. Lett. 74, 1202-1205 (1995). [19] M. Forsthuber and G. Hilscher, Field and geometry dependence of the ac loss in the critical-state model of type-II superconductors. Phys. Rev. B 45, 7996-8006 (1992). [20] V. V. Val’kov and B. P. Khrustalev, Magnetization of granular high-Tc superconductors in strong magnetic fields. JETP 80, 680-685 (1995). [21] T. H. Johansen, M. R. Koblischka, H. Bratsberg, and P. O. Hetland, Critical-state model with a secondary high-field peak in Jc(B). Phys. Rev. B 56, 11273-11278 (1997). [22] D. M. Gokhfeld, Secondary peak on asymmetric magnetization loop of type-II superconductors. J. Supercond. Novel Magn. 26, 281-283 (2013). [23] M. Xu, D. Shi, and R. F. Fox, Generalized critical-state model for hard superconductors. Phys. Rev. B 42, 10773-10776 (1990). [24] D. Dew-Hughes, Flux pinning mechanisms in type-II superconductors. Philos. Mag. 30, 293-305 (1974). [25] E. W. Urban, Flux flow and a new critical-current formula. J. Appl. Phys. 42, 115127 (1971). [26] I. L. Landau, J. B. Willems, J. Hulliger, Detailed magnetization study of superconducting properties of YBa2Cu3O7−x ceramic spheres. J. Phys.: Condens. Matter 20, 095222 (2008). [27] P. Fabbricatore, C. Priano, A. Sciutti, G. Gemme, R. Musenich, R. Parodi, F. Gömöry, and J. R. Thompson, Flux pinning in Bi-2212/Ag-based wires and coils. Phys. Rev. B 54, 12543-12550 (1996). [28] J. Horvat, S. Soltanian, A. V. Pan, and X. L. Wang, Superconducting screening on different length scales in high-quality bulk MgB2 superconductor. J. Appl. Phys. 96, 4342-4351 (2004). [29] M. Eisterer, Calculation of the volume pinning force in MgB2 superconductors. Phys. Rev. B 77, 144524 (2008).




Chapter 8

SUPERCONDUCTING AND MULTIBAND EFFECTS IN FeSe WITH AG ADDITION E. Nazarova1 , N. Balchev1, K. Buchkov1, K. Nenkov2,3, D. Kovacheva4, D. Gajda3 and G. Fuchs2 *

1

Institute of Solid State Physics, Bulgarian Academy of Sciences, Sofia, Bulgaria 2 Leibniz Institute for Solid State and Materials Research, (IFW Dresden), Dresden, Germany 3 International Laboratory of High Magnetic Fields and Low Temperatures, Wroclaw, Poland 4 Institute of General and Inorganic Chemistry, Bulgarian Academy of Sciences, Sofia, Bulgaria

ABSTRACT FeSe crystals and polycrystalline samples with and without Ag addition were prepared by means of different methods. Using transport and magnetic measurements, many important characteristics were investigated in normal and superconducting states. It was established that the Ag addition enhances the critical temperature, the critical magnetic field, Bc2(0), the critical current density, the magnetoresistance, and the pinning energy. The Hall-effect measurements demonstrated that the temperature dependence of the Hall constant, RH(T), is nonlinear and shows a sign-reversal in Ag-doped samples. In consistency with the non-compensated nature of the samples, negative RH values were observed in a large temperature interval; in spite of the higher electron number in these samples, positive (and almost positive) RH values were also detected. The latter underlines the importance of the charge mobility and its large value for holes carriers at higher temperatures. Furthermore, important characteristics were also observed in the magnetoresistance behavior, namely, its increase with the Ag concentration and failure of *

Corresponding Author address Email: [email protected].


196

E. Nazarova, N. Balchev, K. Buchkov et al. Kohler scaling at higher temperatures, as expected in view of the multiband structure of FeSe. However, at temperatures lower than ~30 K, the Kohler plot was restored, indicating the importance of one type of carriers for the magnetotransport. Thus, the Ag addition not only improves the material’s quality and its superconducting properties, but offers also possibilities of investigating the influence of the carriers’ type and concentration on important characteristics of the superconducting FeSe system.

Keywords: FeSe, Ag magnetoresistance

addition,

superconducting

properties,

Hall

effect,

1. INTRODUCTION Explaining the high-temperature superconductivity (HTS) mechanism has been a serious challenge to the condensed-matter physics, ever since Bednorz and Muller discovered superconductivity in the La-Ba-Cu-O compound at Tc ~ 30 K [1]. The advent of iron-based superconductors [2] showed that HTS is not a unique phenomenon. Superconductivity exists in this new class of superconductors even in the presence of a chemical element with a magnetic moment, which is very surprising in the context of the famous Matthias rules. The iron-based superconductors resemble cuprates in what concerns their magnetic properties  for both classes, introducing additional carriers or pressure suppresses the magnetic order and stimulates the appearance of superconductivity. The exact role of magnetism in HTS is yet to be cleared  is it a driving force for superconductivity, is it itinerant with long range or localized and shortrange, or is it simply undesirable? According to one of the leading concepts, the “glue” binding the electrons in Cooper pairs is provided by the exchange of magnetic fluctuations [3]. The spin-fluctuation model predicts many of the normal and superconducting-state properties of cuprates and has been supported by a number of experiments [4]. Another special feature of all iron-based superconductors is their multiband structure. What is usually emphasized in this respect is their similarity to MgB 2 with its two-band structure. However, it should be pointed out that the concept of a multiband structure in cuprates is not to be excluded. The idea for the reconstruction of the electron spectrum by doping [5, 6], the presence of three different Fermi surface arcs in Bi-2212 [7], the explanation of the Hall-coefficient sign-reversal in Y-248 [8] are all important points of discussion concerning the multiband structure in cuprates. Whether this structure is a necessary condition for high-temperature superconductivity is still difficult to affirm. However, the multiband structure leads to many interesting normal-state properties, such as magnetoresistance and Hall-coefficient sign-reversal, so its careful study should yield important information on the necessary conditions for the appearance of high-Tc superconductivity.


Superconducting and Multiband Effects in FeSe with Ag Addition

197

In conventional superconductors, the superconducting pairing and condensation appear simultaneously at Tc. The Cooper pairs strongly overlap in the coherence volume 4πξ3/3 (ξ being the coherence length) as a result of their large number and the region of superconducting fluctuations above Tc is very limited. In high-temperature superconductors (cuprates and iron-based), strong fluctuation effects have been found as a result of preformed Cooper pairs above Tc [9, 10]. In FeSe, this issue is controversial. Both strong superconducting fluctuations [10] and a weak superconducting fluctuation within a very narrow fluctuation region above Tc have been suggested for FeSe [11]. The YBCO system, comprising the first “nitrogen” superconductor, became one of the emblems of cuprates and has been used as a model system for many different theoretical and experimental investigations. The superconducting FeSe has a similar role in the family of iron-based superconductors. With its simple quasi-two-dimensional structure of Fe-Se layers (without a separating layer), and as consisting of two chemical elements only, it offers good opportunities for synthesis and theoretical modeling. The simple tetragonal structure and the experimentally-determined lattice parameters have been used for band-structure calculations [12]. In the framework of density functional calculations, a complicated Fermi surface is found with a cylindrical hole surface centered around the Г (p = 0,0) point, together with cylindrical electron sections around the M point (p = Q = (π, π)) in the Brillouin zone. Cylindrical shapes are open along the c-axis, which are more warped than in cuprates, resulting in a smaller electrical anisotropy in Fe-based compounds. It is very important to underline that the discussed Fermi surface structure concerns the bulk FeSe samples. In a FeSe monolayer on SrTiO3, angle-resolved photoemission spectroscopy (ARPES) experiments have only shown the presence of electronic Fermi surface sheets around the M point, with no hole sheets around the Г point [13]. This reconstructed band structure is different from that of the other Fe-based superconductors and is obviously important for superconductivity and Tc enhancement. Reconstruction of the band structure could be achieved by electron doping [14] or increased pressure [15]. As revealed by band-structure calculations and ARPES studies, the iron-based superconductors are semimetals. In semimetals, the valence and conduction bands overlap slightly without an energy gap between them. Electrons from the previouslyfilled valence band can move to the conduction band, as the highest energy occupied state (the Fermi energy) in both bands is the same. This results in the formation of small electron and hole “pockets” of current carriers and allows a new type of superconducting pairing, known as “s±” pairing [16]. The structural transition in FeSe (from a tetragonal to an orthorhombic crystal structure) established at about ~90 K has also an important effect on the electronic structure. This structural transition is accompanied by the appearance of an electronic nematic state. A nematic state is a form of electronic order with saved translational symmetry of the lattice, but broken rotational symmetries [17]. Various explanations of


198

E. Nazarova, N. Balchev, K. Buchkov et al.

the observed nematicity have been given: (i) generated by spin fluctuations or (ii) resulting from unequal occupancy of the Fe d-electrons’ dxz and dyz orbitals due to the different local environments. This nematic state has an impact on the Fermi surface deformation and, respectively, on the material’s properties. Coming back to the FeSe investigated here, we should point out that the stoichiometric compound has an equal number of electrons and holes [18]. However, this balance is disturbed in the Se-deficient iron selenide (FeSe0.94) investigated. Incorporating Ag into the crystal lattice unit cell could increase this imbalance further, causing important changes in many properties. Below we will summarize our detailed investigations and present new results on the FeSe system and the effect of adding Ag on many important parameters. We found that Ag improves the material’s quality and its superconducting properties (increases Tc, Bc2, Hirr, Jc, k – Ginzburg-Landau parameter, pinning energy). Moreover, the Ag addition allows one to investigate the effect of the carriers’ type and concentration on important normal-state characteristics of the superconducting FeSe system.

2. SYNTHESIS AND STRUCTURE We synthesized polycrystalline FeSe0.94 samples by a solid-state reaction and partial melting. This was a two-stage process with intermediate grinding and heating up to 700°C and 1050°C, respectively. Crystals of Fe1.02Se were also grown following the NaCl/KCl flux technique from pre-sintered powders heat-treated at 850°C [19]. All procedures were carried out in a glove box filled with Ar, with the synthesis performed in evacuated (10-3 Torr) and sealed quartz ampoules [20]. In the case of melted polycrystalline samples and crystals, and due to the growth process at high temperatures (>750°C) when the hexagonal phase is more stable [21], this phase was always present in the final samples. This result was reached in spite of annealing the samples for a long time (~100 h) at 400°C (which was found to reduce the hexagonal phase amount) and using the composition FeSe0.94, where this amount should be minimized (b>c) FeSeAg6 sample using a Quantum Design PPMS-14. The magnetic field was applied perpendicular to the sample surface (a x b), with its direction being changed during the Hall measurements. Four contacts were formed on the sample surface along the x axis (sample’s “a” dimension was parallel to the x axis) – two of them for current leads and the other two for the magnetoresistance signal denoted below as ρxx. Two other contacts were placed along the y axis, which was parallel to the sample’s “b” dimension. The signal measured on these contacts is denoted as ρxy; the Hall resistance was found according to the relation ρH = [ρxy(+B) - ρxy(-B)]/2, where ρxy(+B) and ρxy(-B) are the resistances measured along the y axis in the two opposite directions of the magnetic field. In non-magnetic materials, the Hall effect represents the appearance of a transverse potential difference in the magnetic field due to the Lorenz force acting on the current carriers. However, a supplementary part proportional to the magnetization appears in magnetic materials, known as anomalous Hall effect (AHE); the Hall resistivity is then [25]:

 H  R0 H  Rs M ,

(1)

where R0 and Rs are the normal and anomalous Hall coefficients, respectively, H is the applied magnetic field and M is the magnetization. The anomalous Hall effect is due to the spin-orbit coupling in the presence of spin polarization [25]. Our crystals, as obtained by the flux technique, consisted of about 18% of a hexagonal magnetic phase [19], which was expected to generate a significant AHE. Recently, single crystals of very good quality have been obtained by the vapor-transport technique under temperature-gradient conditions [26, 27]. However, the single-crystal stoichiometry results in an almost compensated sample with similar numbers of electrons and holes. The FeSe0.94 samples with Ag additions do not contain a hexagonal phase (or its amount is below the XRD analysis sensitivity) and, unlike the single crystals, are strongly non-compensated (with a higher number of electron carriers). This gives one unique possibilities to investigate the magnetotransport in non-compensated samples. To study the AHE in our FeSeAg6 sample, we examined the magnetic field dependence of the transverse resistivity (ρxy)



201

with a view to finding the field range where the AHE is significantly smaller than the normal one. Figure 2 shows the field dependences of: (a) the transverse resistivity ρxy(B) and (b) the Hall resistivity ρH = [ρxy(B) - ρxy(-B)]/2 at T = 300 K of FeSeAg6-#1 sample. As one can see in Figure 2b, a step-like behavior, indicating the presence of AHE, is observed in the interval -2.5 T < B < 2.5 T. In Figure 3 we show the M-H curve of the sample investigated. As seen, the magnetization saturates above and below the same field interval (-2.5 T < B < 2.5 T). We established that this saturation persisted in a wide temperature interval, namely, 2 K ≤ T ≤ 300 K. The saturation magnetization Ms was about 20 emu/cm3. This value is close to that obtained by the authors of [28] for FeSe thin films. The low value of Ms obtained suggested that the AHE could be neglected. To check for this point, we obtained RH by fitting linearly ρH (Figure 2b) twice: at B ≥ 0 (neglecting the slope change around 2.5 T) and at B ≥ 2.5 T. We thus obtained: RH = 2.51×10-7 m3/C in the former case, and RH = 2.03×10-7 m3/C in the latter.

Figure 2. Field dependence of: (a) – the transverse resistivity ρxy and (b) – the Hall resistivity ρH at T = 300 K of the FeSeAg6-№1 sample.



202

Figure 3. Field dependence of the magnetization for Ag doped sample at 2 K.

The close values for RH in the two cases confirm again the small AHE. According to relation (1), the Hall resistivity for FeSe is determined mainly by the first term, which increases as the field is increased, while the second term remains constant at fields larger than 2.5 T. It is important to stress that the positive value of RH means that the hole-type carriers are dominant at T = 300 K in FeSeAg6-#1. The Hall effect is used not only to determine the carrier type, but its concentration as well. In the case of materials with only one type of carriers (electrons or holes), this is easily done by using the relation RH = -1/nee, where ne is the electron concentration, and RH = 1/nhe, with nh being the hole concentration. However, in multiband materials (as FeSe) and semiconductors, both carriers’ types exist and take part in the transport simultaneously. In this case, the mathematical expression for the Hall constant is more complicated [29]:

(2) and reduces to:

RH  (h nh  e ne ) / e(e nh  h ne ) 2 2

2

(3)

in the limit of low fields H →0 and

RH  1 / e(nh  ne ) in the limit of large fields.


(4)


203

Here ne and nh are the electron and hole concentrations; and μe and μh are the electron and hole mobilities. As (3) indicates, the sign of the Hall coefficient is positive if (μh2nh – neμe2) > 0 and negative in the opposite case. Thus, it appears that at low fields the Hall coefficient is affected to a larger extent by the mobility of the charge carriers than by their number density in the sample. This is why the exact carrier concentration can only be determined at low temperatures and very high magnetic fields [30]. Assuming that the slope of ρH vs B does not change at higher fields and using (4) and Figure 2b, one can calculate the effective carrier concentration nH = nh – ne = 1/eRH. The values of nH thus obtained were: 2.49 × 1019 cm-3 at B ≥ 0 and 3.08 × 1019 cm-3 at B ≥ 2.5 T, respectively, and T = 300 K. One could, therefore, conclude that the AHE does not affect significantly the carrier concentration of the investigated sample. Since the Ag-doped sample studied might contain traces of Ag2Se [22], we could compare the value obtained of RH with that for Ag2Se. The Hall coefficient of β-Ag2Se has been reported to be (4-7) × 103 cm3/C at high temperatures [31]. This would correspond to a carrier concentration of ~10 15 cm-3, which is significantly lower than the one obtained for our sample. For a normal metal with a Fermi-liquid behavior, the Hall coefficient as a function of the temperature is constant [32]. However, the Hall coefficient varies with the temperature for multiband samples (such as MgB 2 [33]), or samples with a non-Fermi liquid behavior (such as cuprate superconductors [34]). In the case of the FeSeAg6 superconductor reported here, RH varied with the temperature. Did the temperature dependences of RH observed by us in polycrystalline samples show an intrinsic property of iron-based superconducting FeSe? The nonlinear temperature dependence exhibited by single-crystal FeSe [18, 35-37] suggests a positive answer. We should also add that a multiband structure [12] and a non-Fermi liquid behavior [20, 22] were characteristic for all our samples. Figure 4(a, b) presents the temperature dependences of the Hall coefficient, RH, of two different pieces (#1 and #2) cut from sample FeSeAg6. The averaged curves were plotted using two separate temperature scans under a fixed magnetic field B = ±1 T. In Figure 4a, a sign reversal of RH is seen at about 240 K. For the other sample (Figure 4b), slightly positive values of RH at about 225 K are present. This temperature discrepancy results from the polycrystalline nature of the samples, their layered FeSe structure, and the non-homogeneous Ag distribution. Let us mention two important facts: (i) as consistent with the non-compensated nature of the samples, negative RH values were observed in a large temperature interval, and (ii) despite the higher electron number in these samples, positive (or nearly positive) RH values were also detected. The latter underlines the importance of the charge mobility, i.e., its large value for holes carriers at higher temperatures.


204


Figure 4. Temperature dependences of the Hall coefficient RH of Ag-doped sample.

A similar sign reversal of RH has been observed in FeSe thin films at about 185 K [28]. These temperatures are significantly higher than that of the structural phase transition (about 90 K). Moreover, the latter is not influenced by the Ag doping [22]. Therefore, this phenomenon is not due to a structural phase transition. The sign reversal is not influenced by the sign of the anomalous Hall term [38]; it indicates that the material has hole-like carriers at high temperatures, and then changes to electron-like ones at lower temperatures. This can be explained bearing in mind the different values of the hole and electron mobilities, μh and μe, at the different temperatures. At high temperatures μh > μe, meaning that the hole channel dominates the normal-state transport. At low temperatures, the hole channel can be replaced by the electron channel, as in the case of FeTe [38]. The authors of [39] observed a remarkable reduction in the carrier number and an enhancement in the carrier mobility below 120 K in FeSe single crystals. The mobility spectrum implied the presence of a minority of ultrafast electron-like carriers. Recently, Watson et al. [18] interpreted this result on the basis of a three-band model – in addition to one hole and one almost compensated electron band, the orthorhombic phase of FeSe exhibits an additional tiny electron pocket of high mobility.



205

According to relation (2), the Hall coefficient’s magnetic field dependence at different temperatures should be quadratic for two-band structure materials. We performed a detailed examination of the ρxy(H) dependences at different temperatures, starting from T = 15 K up to 100 K with a step of 10 K, and up to T = 250 K with a step of 20 K or 50 K. These experiments showed that at low temperatures ρxy(H) is almost linear, but changes to a quadratic behavior at higher temperatures. At selected temperatures, both ρxy(+H) and ρxy(-H) were measured; the ρH(H) dependences thus obtained are presented in Figure 5. At low temperatures (15 K – 40 K), the ρH(H) dependences are negative and almost linear when the field is increased. This is in agreement with the ρH(T) (presented in Figure 4b) when the electron carriers are dominant. The sample behavior under these conditions resembles that of materials with only one type of carriers, which persists to temperatures ~225 K  240 K. The strong negative signal for the ρH(T) dependences was detected at low temperatures, demonstrating the presence of a high number of electron carriers in this non-compensated Ag-doped FeSe specimen. As the temperature was increased and the point of sign change thus approached, the product μini (i = h,e) was balanced and both types of carriers became equally active. At these temperatures, the ρH(H) dependences strongly deviated from the linear behavior, demonstrating a multiband structure and two types of carriers. The experiment provided evidence that, in spite of the large number of electrons, both type of carriers become active as a result of the higher holes mobility.

Figure 5. Magnetic field dependence of Hall resistance at several selected temperatures for FeSeAg6-№2.


206


3.2. Magnetoresistance The magnetoresistance effect ([ρxx(H) – ρxx(0)]/ρxx(0) = Δρ/ρ0) was registered at 11 K and a magnetic field B = 12 T for all Ag-doped samples, as well as for the non-doped one. The results obtained are presented in Table 1. Table 1. Residual resistivity ratio (RRR); Magnetoresistance (MR) Δρ/ρ0 at 11 K and 12 T; Tc at ρ = 0 and B = 0, Transition width (ΔT) at B = 0 according to 10-90% criterion Sample FeSeAg0 FeSeAg4 FeSeAg6 FeSeAg8 FeSeAg10

RRR ρ(300)/ρ(16) 2.95 9.55 9.74 9.47 3.25

MR (%) 6.4 50.5 36.4 39.6 34.3

Tc(ρ = 0, B = 0), (K) 7.30 8.78 9.05 8.60 8.10

ΔT(B = 0), (K) 2.27 0.85 0.76 0.70 0.93

It is seen that the non-doped sample exhibits a negligible magnetoresistance (MR) effect, while the Ag addition in the other samples enhances significantly this effect. Zero MR has been predicted for a single-band free-electron system [40]. In the lowfield limit ωcτ 80% of the theoretical value, and high values of the transport critical current density mainly at grain boundaries. Over the past years, the increasing use of nonconventional electric current activated/pressure assisted densification methods as a tool for powder densification has demonstrated its effectiveness in obtaining highly dense and textured ceramic materials. This densification technique that involves the combined application of electrical current and uniaxial pressure to a powder sample is known commercially as Spark-Plasma Sintering (SPS). In this Chapter, we present a review of the application of SPS into consolidate ceramics samples of (Bi-Pb)2Sr2 Ca2 Cu3O10+δ (Bi-2223) superconductors. A series of studies performed under different SPS conditions such as consolidation temperatures and consolidation times, uniaxial compacting pressures, and die material are discussed. According to the literature, the SPS consolidation of the Bi-2223 superconductor conducted in vacuum was found to result in samples with relative densities of 90%, even when a low compacting pressure of 50 MPa is used. The consolidation process performed in vacuum drastically alters not only the surface of the grains, but also their grain boundaries, leading to the occurrence of grains with a core-shell structure and comprised of: (i) a core with the stoichiometric Bi-2223 phase; and (ii) an oxygen-deficient shell. Such an oxygen deficiency at the surface of the grains was found to be responsible for the suppression of tunneling of Cooper pairs between adjacent superconducting grains and therefore a drastic reduction of the transport superconducting current density of the SPS materials. In order to restore the oxygen content of the SPS samples, a post-annealing heat treatment (PAHT) performed in air, at selected temperatures, and for a brief time interval of 5 min is usually needed. The PAHT results in a decrease in the width of the oxygen-deficient shell and an increase in the oxygen content along the grain boundaries, triggering the formation of conduction current paths along the grain boundaries of the dense material. SPS samples subjected to the PATH exhibit a 10 to 25-fold increase in their transport superconducting current density at 77 K. In order to shed light in the understanding of the relationship between the consolidation conditions, the microstructure, and the physical properties of the SPS samples, a finite element simulation method has been used and compared with experimental results. The most important factors to be considered for practical applications of the SPS Bi-2223 materials are also discussed. The chapter is organized as follows: (1) Introduction; (2) Experimental Procedures; (3) Finite Element Simulations; (4) Results and Discussion; and (5) Final Remarks.

Keywords: bi-based superconductor, spark plasma sintering, finite element method, superconducting properties

1.

Introduction

In cooper-oxide superconducting ceramics the transport properties are mostly determined by its granular nature. The presence of current-path frustration such as cracks, voids, and grain boundaries constitutes an unavoidable fact due to a different conventional obtention process from powdered precursors. The particular case of the grain boundary plays an important role in limiting the transport properties of these materials. It is believed that this


Recent Progress in Powder Densification of (Bi,Pb)-Sr-Ca-Cu-O Materials

227

phenomenon is mainly due to both the misorientation and the electrical connectivity between grains [1]. High angle grain boundaries can act as Josephson coupled weak links, leading to a significant field-dependent suppression of the supercurrent across the grain boundary [2]. Therefore, the manufacture of highly dense and textured cooper-oxide superconducting ceramics is a matter of great importance in producing materials with optimal transport properties for technological applications. In this scenario, the uniaxial and the isostatic compacting pressure are two of the most widely used methods for inducing powder densification. These techniques, also referred to as powder pressing [3], are intended to manufacture net-shaped and homogeneously dense pellets expected to be free of defects. Since the discovery of high-Tc superconductors, several studies have been devoted to disclosing the influence of these types of mechanical deformation on their general transport properties. These studies have frequently indicated that increasing the uniaxial compacting pressure results in samples with either a higher volume density and texture degree along with an improvement in the intergrain connectivity [4]. Also, the obtained results have shown an increase in the pinning force and consequently in the transport critical current density across grain boundaries. However, depending on the mechanical properties of the processed powder, the uniaxial compacting pressure may result in defects such as end capping, ring capping, lamination, voids, and vertical cracks [3, 5]. These defects limit the achievement of high volume density greater than 80% of the theoretical value, and high values of the transport critical current density mainly at grain boundaries. Over the past years, the increasing use of nonconventional electric current activated/pressure assisted densification methods as a tool for powder densification has demonstrated its effectiveness in obtaining highly dense and textured ceramic materials. This densification technique which involves the combined application of electrical current and uniaxial pressure to a powder sample, is known commercially as Spark-Plasma Sintering (SPS). The application of the SPS results in materials with remarkable physical properties [6], possibly associated with the reduction of the impurity segregation at grain boundaries and relative densification, is frequently higher than 90% of the theoretical value [8]. In addition, the benefits of the method include lower sintering temperatures and high heating rates [7]. In this chapter, we present a review of the application of SPS into consolidated ceramics samples of (Bi-Pb)2 Sr2 Ca2 Cu3 O10+δ (Bi-2223) superconductors. A series of studies performed under different SPS conditions and the most important factors to be considered for possible practical applications of the SPS Bi-2223 materials are discussed.

2.

Experimental Procedures

Powder samples of Bi1.65 Pb0.35 Sr2 Ca2 Cu3 O10+δ ((Bi,Pb)-2223) were obtained by using the traditional solid-state reaction method from Bi2 O3 , PbO, SrCO3 , CaCO3 , and CuO, which were mixed in an atomic ratio of Pb:Bi:Sr:Ca:Cu (0.35:1.65:2:2:3). The mixture was first calcined in air at 750 ◦ C for 2400 min. Then, the powder was reground and pressed into pellets at a pressure of 250 MPa. These pellets were heat treated at 800 ◦ C in air for 2400 min. Afterward, the pellets were reground, pressed again, and sintered in air at 843 ◦ C for 2400 min in a Lindberg/Blue tubular furnace. This step was repeated three times.


228

L. Pérez-Acosta, E. Govea-Alcaide, R. de Figueiredo Jardim et al.

After the last heat treatment, the pellets were reground again and the resulting powder was consolidated by spark plasma sintering (SPS). For the SPS consolidation, a small portion of (Bi,Pb)-2223 powders (∼ 4 g) was deposited inside a cylindrical die made of high density graphite with dimensions of 50 mm (outer diameter) and 20 mm (inner diameter) and between two graphite plungers of 40 mm in height. In order to facilitate the sample release after consolidation, graphite foils were inserted between the internal surfaces of the die and between the top and the bottom surfaces of the sample and the graphite plungers. Also, to determine the influence of spark plasma consolidation on the superconducting properties in other type of material for plungers/die (the AISI H13 steel) was used to consolidate the pre-reacted powders of (Bi,Pb)-2223 with the same dimensions of the graphite plungers/dies. The dies were then placed inside the chamber of the Spark-Plasma Sintering System (SPSS) 1050 Dr Sinterr apparatus, manufactured by Sumitomo Coal Mining Co. Ltd., Japan. This equipment combines a maximum uniaxial pressure of 100 kN with dc pulsed currents up to 5000 A. An important aspect to taking into account in this process is that the sintering occur under high vacuum (10 ∼ 30 Pa). The consolidation process was developed under different sintering conditions. From now on, the following fundamental consolidation parameters will be employed: TD the consolidation temperature, HR the heating rate, tR the heating time, and tD the consolidation dwell time, as the fundamental consolidation parameters employed in this work. The temperature was measured during the consolidation process by using a (K-type) thermocouple, which was inserted into a small hole in one side of the graphite die. For comparison reasons, ∼ 4 g of the starting powder was cold pressed inside the SPS apparatus and the resulting pellet was sintered by the traditional ceramic method, at 845 ◦ C in air for 2400 min. This sample will thereafter be referred as the reference sample (REF). The phase identification of the samples was evaluated from X-ray diffraction patterns obtained in a Bruker-AXS D8 Advance diffractometer. These measurements were performed at room temperature using CuKα radiation in the 3 ◦ ≤ 2θ ≤ 80◦ range with a 0.05◦ (2θ) step size, and 5 s counting time. The volume density, ρv , of all pellets was determined by using the Archimedes method. Three types of magnetic measurements were performed by using a commercial Quantum Design SQUID magnetometer. The first one, the magnetization as a function of the applied magnetic field, M(H), was performed in pellet and powder samples. The M(H) dependence, was taken under zero field cooled (ZFC) conditions, for different temperatures in the range 2 - 95 K in steps of ∼ 15 K. Once the temperature was reached, H was applied perpendicular to the compacting direction of the pellets, and increased from 0 to 15 kOe. Magnetization as a function of temperature, M(T), which was performed under both zerofield-cooled (ZFC) and field-cooled (FC) conditions. The ZFC cycle was performed after cooling down the sample to 10 K without the application of any magnetic field. Then, a magnetic field of H = 50 Oe was applied, and the data were collected during the warming process from 10 to 150 K. Subsequently, the FC measurements were performed by cooling the sample slowly from 150 K down to 10 K in the same applied magnetic field. Finally, the third one is the measurement of the magnetic relaxation curves, M(t), in pellets, for selected temperatures between 2 and 15 K. The time dependence of M was typically recorded over 2 h.



229

Various types of transport measurements were realized. The temperature dependence of the electrical resistivity, ρ(T ), was measured by using the standard dc four-probe technique, and cooling the samples in zero applied magnetic field. The ρ(T ) measurements were performed in the temperature range 30 ≤ T ≤ 300 K, and in the applied magnetic field, H, ranging from 0 to 500 Oe in a closed cycle cryogenic refrigerator ARS-4HW/DE-202N attached to a temperature controller Lakeshore model 331S. For these measurements, Au electrical leads were attached to Ag film contact pads on samples using Ag epoxy. The magnetic field H was always applied perpendicularly to both the thickness of the samples and to the excitation current, I, which has been injected along the major length of samples. The excitation current used was 1 mA. In addition, current-voltage (I − V ) measurements were performed after cooling the sample in zero applied magnetic field to T = 77 K. Once the temperature was stabilized, the excitation current through the sample was applied and increased automatically in steps of 1 mA, while the voltage across the sample was measured. The value of the transport critical current density at zero applied magnetic field, Jc , was determined from the measured I − V curve by taking the Jc value when the voltage across the sample reaches 1 µV. Also, thermal conductivity measurements were carried out in the temperature range of 2 - 300 K using the thermal transport option (TTO) of the Quantum Design PPMS system. Finally, all SPS samples were subjected to an additional post-annealing heat treatment (PAHT) performed in a tubular furnace, in air, at different temperatures and times. Sample nomenclature, some SPS conditions, and other important parameters are shown in Table 1. Table 1. Consolidation parameters used during the SPS process for producing Bi-2223 samples. TD is the consolidation temperature, HR is the heating rate, tr is the heating time, and tD is the consolidation time. We also included other important parameters (see the text for details) Sample S750 S800 S830 REF P750 P800 P830 H700 H750

3.

TD (◦ C) 750 800 830 845 750 800 830 700 750

HR (◦ C/min) 145 155 160 5 145 155 160 135 50

tr (min) 5 5 5 163 5 5 5 5 15

tD (min) 5 5 5 2400 5 5 5 5 5

ρv (g/cm3 ) 5.7 5.6 5.6 3.2 5.7 5.6 5.6 4.8 5.5

ke (300K) (K) 0.07 0.02 0.08 -

Jc(77K) (A/cm2 ) 10.0 5.4 2.1 21.8 128.2 87.2 58.1 -

B0 (mT) 0.64 0.42 1.02 -

ρ(300K) (mΩcm) 10.0 15.8 32.6 9.3 4.3 5.3 6.8 4.3 9.0

Finite Element Simulations

The SPS setup used for processing our samples does not allow simultaneous measurements of temperature and electric current intensity, during the sample consolidation. For this reason, it is necessary to execute a procedure in order to determine the values of these parameters. As the finite element method (FEM) in the SPS electro-thermal problem model


230


is a powerful tool for comparing experimental and theoretical data of SPS processes, it has been employed here for our analysis [9]. The arrangement consists of two Inconel electrodes, six graphite spacers, the die (thickness 11 mm, height 21 mm) with two plungers (diameter 26 mm, height 20 mm) surrounding the sample (diameter 26 mm, height 2 mm), the latter located in the center of the stack. Taking into account the axisymmetrical configuration, only two dimensions are of interest: the system is studied in cylindrical coordinates and the problem is rather simplified. The schematic drawing of the consolidation system and boundary conditions are displayed in Figure 1(a). For each one of the above domains, the electro-thermal process is described by two coupled partial differential equations: one related to the charge conservation law and the other one associated with the heat transfer process by itself [10, 11] ∇ · J = ∇ · (σE) = ∇ · (−σ∇V ) = 0,

(1)

ρc p (T )∂T /∂t − ∇ · (k∇T ) = q˙J .

(2)

where T is the temperature, ρ is the density, assumed to be constant and corresponding to the last stage of sintering, q˙J = JE is the heat loss by Joule heating per unit volume per unit time, J = E/ρe(T ) is the electric current density, ρe(T ) is the temperature dependence of the electrical resistivity, c p (T ) is the heat capacity as a function of temperature, and k(T ) is the thermal conductivity. Values for each parameter are assigned for each material or domain in the geometry, i.e., the Inconel 600, the high-density graphite, the AISI H13 steel, and the Bi-2223 sample. Table 2 displays the parameters used in the FEM simulations.

Figure 1. (a) Schematic drawing of the consolidation system, (b)boundary conditions, and (c) simulation by FEM of the temperature distribution of the whole system.



231

The initial and boundary conditions used for solving Eqs. (1) and (2) are displayed in Figure 1(b). The initial temperature was set to be 300 K and the heat losses by conduction and/or convection through the gas were neglected because the process occurs in vacuum. All the free surfaces exposed to the vacuum chamber have heat losses by radiation given by q˙rad = σs ε(Td4 − T04 ), where Td is the temperature of the free surfaces, σs the StefanBoltzmann’s constant, ε = 0.3 (0.69) is the graphite (Inconel 600) emissivity, and T0 = 300 K is the temperature of the chamber wall. The temperature of both the upper and the lower Inconel electrodes was 300 K and the electrical and the thermal resistances were disregarded [10]. A preset time-current profile I(t) is directly applied on the top of the electrode. Finally, the electro-thermal Eqs. (1) and (2) were solved by using the COMSOL MultiphysicsTM package. Table 2. Parameters of Bi-2223 samples, graphite, Inconel 600, and the H13 steel used in the FEM simulations: ρ is its density, ρe(T ) is the electrical resistivity as a function of temperature, c p (T ) is the heat capacity as a function of temperature, and k(T ) is the thermal conductivity Physical Property ρ ρe(T )

Bi-2223 5700∗ 0.8 × 10−6 + 0.8 × 10−8 T

c p (T )

131.6 + 0.77T

k(T)

0.27 + 1.95 × 10−3 T

High-density graphite 1900 2.4 × 10−5 − 2.6 × 10−8 T + 2.2 × 10−11T 2 −151 + 2.3T − 7.1 × 10−4T 2 84 − 0.063 × T + 2.9 × 10−5 × T 2

Inconel 600 8175 1.03 × 10−6 − 1.85 × 10−10T + 6.2 × 10−13T 2 395.28 + 0.1576T + 7.1 × 10−5 T 2 8.955 + 0.016T

AISI H13 7670 4.3 × 10−7 + 1.7 × 10−10T + 5 × 10−13T 2 488.044 − 0.2T + 3 × 10−4 T 2 −11.36 + 0.15T

Units kg/m3 Ωm

J/kg K

W/m K

There are two forms of controlling the SPS process: (i) temperature control; and (ii) current intensity control. In our simulations, the former has been employed, i.e., a preset time-current profile I(t) is directly applied as a function of the measured temperature in a close-loop control system in such a way that a preset time-temperature T (t) profile is reached. A proportional-integral-derivative (PID) control has been programmed into the COMSOL code. The current profile I(t) is given by I(t) = kP e(t) + kI

Z t 0

e(t)dt + kD (de(t)/dt),

(3)

where e(t) is the difference between the desired temperature and the actual temperature of the control point. Values of kP , kI , and kD are then obtained by adjusting the PID control with the Ziegler-Nichols method [12].


232

4. 4.1.


Results and Discussion Structural and Phase Identification

As a general rule, the X-ray diffraction technique constitutes a powerful tool to probe and understand the structure and to obtain atomic spacing of crystalline compounds. It provides information about structure, phases, average grain size, texture (preferred crystal orientations), as well as many other microstructural parameters. This common technique is based on constructive interference produced by the interaction between X-rays and crystalline samples, when Bragg’s law 2dsinθ = mλ is satisfied. Here, d is the lattice spacing in a crystalline sample, θ is the diffraction angle, m is the diffraction order and take integer values, and λ is the wavelength of the X-rays. All possible diffraction directions of the lattice should be reached by scanning over the sample for an interval of 2θ angles, due to the random orientation of the crystallites in powder samples. Then, the diffraction peaks are converted to d-spacings that allows identification of the compound because each compound has a set of unique d-spacings. Generally, this is achieved by comparing of d-spacings with reference patterns. In our case, the microstructure analysis by X-ray diffraction was performed on powders and pellets of the reference and the SPS samples, and the patterns are displayed in Figure 2. The results reveals that all indexed reflections are related to the high-Tc Bi-2223 phase. The peaks of the (Bi,Pb)-2223 phase were then indexed with respect to an orthorhombic crystal structure, space group A2aa, lattice parameters a = 21.640 Å, b = 5.413 Å, and c = 37.152 Å, in excellent agreement with those reported for the same compound elsewhere [4,13]. By a careful inspection of the data, we also detected an unknown reflection near 2θ = 31.4◦ , not associated with the Bi-2223 phase, in the X-ray diagram of all SPS samples. Such an extra phase was identified as an infinite layer compound with general formula Ca1−x Srx CuO2 [14]. The volume fraction of this extra phase was found to remain essentially constant in all consolidated samples. The kinetic formation of this compound, as a result of the SPS process, still remains unclear.

4.2.

The Influence of Spark Plasma Consolidation Temperature, TD

In order to achieve a better understanding of the SPS process, the role of the consolidation temperature, TD , on the superconducting properties was investigated. It is important to mention that the formation of a single phase (Bi,Pb)-2223 is extremely difficult since it usually interweaves with the (Bi,Pb)-2212 phase. This is due to the high complexity of the reaction and to the small difference in their thermodynamic stabilities [15]. For this reason, the consolidation temperature is one of the most important parameters to take into account in the SPS process. The effect of TD on the magnetic and transport was studied by sintering samples at three different temperatures TD = 750, 800, and 830 ◦ C (see Table 1). We start our analysis with the temperature dependence of dc magnetization, which was measured in powder samples, and the results are displayed in Figure 3. The results can be summarized as follows: (i) Powders of the reference sample REF exhibit higher diamagnetic signals than SPS samples. (ii) The diamagnetic signals of the SPS samples decreases with increasing consolidation temperatures, TD . (iii) The onset of the diamagnetic signals, Tcg , is below the superconducting critical temperature of the grains Tcg ∼ 109, 107, and 103



233

Figure 2. X-ray diffraction patterns of powder of the reference sample REF(a) and the SPS samples S750 (b), (c), and (d). All peaks belong to the Bi-2223 phase. K, for samples REF, S750, and S830, respectively. That is, an increase in the consolidation temperature, TD , results in a systematic decrease of Tcg . (iv) The difference between the ZFC and the FC curves at T = 5 K, ∆M = MFC (5K) - MZFC (5K), which is proportional to the superconducting critical current, indicates that increasing in TD results in a decrease of ∆M. Taking into account that the critical parameter Tcg is oxygen-content dependent [16] and that the SPS process occurs at high vacuum, one may suggest that samples produced by the SPS technique are oxygen-deficient, and, for this reason, display Tcg a bit smaller than the reference sample REF. Thus, higher consolidation temperatures result in higher oxygendeficient samples. Also, as was previously reported [17] in YBa2 Cu3 O6+δ nanopowders, the deoxygenation on the surface of the grains leads to a decrease of the superconducting volume fraction and the diamagnetic signal. Therefore, the grains are believed to have a shell-core morphology in which the shell (surface) is oxygen-deficient. This, in fact, results in a much smaller superconducting volume fraction of the grains, as can be seen from the features of the M(T) curves. As far as point (iv) is concerned, the deoxygenated grains are also responsible for the observed decrease in ∆M with increasing TD . It seems that increasing TD results in a thicker oxygen-deficient surface of the grains, further indicating that the magnetic flux is not trapped within the grains. This occurs because the superconducting core size may be similar or smaller than the London penetration depth λL [17]. The above feature has its counterpart in the magnetic field dependence of the dc magnetic susceptibility, χ(H) = M/H, and can be easily monitored through of the lower critical field values within the framework of the Bean critical state model. For the determination of the lower magnetic critical field, a relevant question concerns the procedure to obtain the magnetic field at which the M(H) curves start to deviate from linearity, which is a stan-


234


Figure 3. M(T) curves in powder samples of REF, S750, and S830 under ZFC and FC conditions and under an applied magnetic field of 50 Oe. The solid lines represents guide for the eyes. dard procedure to determinate this critical parameter. Usually, such value is determined by visual localization, but such a procedure, as has been discussed by some authors, is too subjective to be quantitative. Here, the used procedure was described elsewhere [18]. Figure 4 displays typical χ(H) curves measured in pellets samples of REF, S750, and S830 at 2 K. It is clear that the behavior of all curves strongly depends on the magnitude of H. The χ(H) dependence of reference sample exhibits an intergranular feature for H < Hc1g (see Figure 4), where Hc1g is the first critical field of the grains. First, χ is mostly magnetic field independent up to Hc1 j = 4 mT. It is important to point out that the low-field plateau of χ(H) observed represents the intergranular magnetic flux shielding and Hc1 j is identified as the first intergranular critical field of the material. In the magnetic-field range 4 - 30 mT, χ(H) decreases appreciably because the magnetic flux penetrates the intergranular medium of the material. The increase in the magnetic field, H > Hc1g, results in the penetration of the magnetic flux within grains and the magnetic susceptibility χ(H) decreases further and gradually; Hc1g , displayed in the Figure 4, denotes the first magnetic critical field of the superconducting grains. The procedure was applied for the remainder samples as well. For the SPS samples, Hc1 j reached the value of 1 and 2.5 mT for S830 and S750 respectively, which are lower values than those obtained for the reference sample. That is, even if the grains are tightly packed in the SPS samples, other current path frustration mechanisms have an important contribution in the shielding currents due to the oxygen-deficient surface of the grains. In a simplified manner, we can say that the increase in the consolidation temperature, TD , leads to a weakening of the superconducting properties at the grain-boundary region.



235

Figure 4. χ(H) curves for pellet of samples REF, S750, and S830 at T = 2 K. The solid lines represents guide for the eyes. It is reasonable to point out that the grains morphological alterations provoked by increasing the consolidation temperature and the effects of the deoxygenation must be reflected in other macroscopic properties of these SPS samples. Specifically, grain boundaries play an important role in limiting the general transport properties of polycrystalline superconductors [2]. Following this statement, the temperature dependence of the electrical resistivity, ρ(T), of samples REF, S750, S800 and S830 are displayed in Figure 5. The ρ(T) curves of the reference sample REF exhibit a transition to the superconducting state below the onset critical temperature Ton ∼ 114 K. The temperature in which ρ(T) = 0 is observed, To f f , reaches a value of ∼ 103 K. Due to the prolonged sintering time, the grains of the reference sample are homogeneous and not oxygen deficient, being responsible for high values of Ton . In the samples SPS, Ton takes the values 110, 108.2, and 106.3 K and To f f ∼ 83, 83, and 81 K for S750, S800, and S830 respectively. These results indicate that the higher the consolidation temperature TD is, the higher is the oxygen depletion of the samples, and the lower is Ton . We also found that To f f increases slightly from 81 K, in sample S830, to 83 K in samples S800 and S750, further indicating that grain boundaries become cleaner with decreasing TD . In the normal-state region (T > Ton ), the ρ(T) curve of the reference sample display a typical T -linear behavior. However, the ρ(T) curves of the SPS samples are quite different from the one of the reference sample. The data shows that sample S830 exhibits a semiconductor-like behavior down to ∼ 110 K, and reaches a maximum near the onset critical temperature. Its value of ρ(300K) is as high as ∼ 33 mΩcm. The ρ(T) behavior of sample S800 is quite different and ρ(T) is essentially temperature independent in a wide range of T (100-300 K), exhibiting similar behavior as in disordered metals. We also men-


236


tion that ρ(300K) ∼ 16 mΩcm for the SPS sample S800, a value two times lower than the one found in sample S830. On the other hand, sample S750 exhibits a typical metallic behavior with ρ(300K) ∼ 10 mΩcm, a value close to the one of the reference sample REF (see Table 1). The inset of Figure 5 displays features of the superconducting transition which must also be considered. Here, the important parameters are Ton and To f f , and consequently the transition width. With the progressive decrease of TD the onset critical temperature Ton increases from ∼ 106 K to ∼ 110 K between samples S830 and S750, respectively. The general behavior of the ρ(T) curves indicates that by decreasing the consolidation temperature TD both the normal and superconducting transport properties of the samples improve. Such a behavior is strongly related to the oxygen content of the samples and the width of deoxygenated-shell decreases with decreasing TD [19]. Another important information regarding the effects of the consolidation temperature on the transport properties can be extracted from the temperature dependence of the thermal conductivity, k(T). Figure 6 displays the k(T) curves for reference sample REF and the SPS samples S750 and S830. The total thermal conductivity consists of the phonon and the electronic contribution, which is connected to the electrical conductivity, in first approximation, by the Wiedemann-Franz law. In all samples, k(T) increases with increasing temperatures between 2 to 300 K, and SPS samples were found to display higher values of k(T) when compared with the reference sample, REF, a feature that is more pronounced in the normal-state region (T > Tck ). This behavior of k(T) is certainly related to the higher volume density of SPS samples, which displays a very low concentration of voids. The electronic contribution to ke (T) was estimated by using the Wiedemann-Franz law, ke (T)=L0 T /ρ(T ), where L0 = 2.45 x 10−8 WΩ/K2 is the Lorentz number. Note that all values of ke , which are listed in Table 1, are at least one order of magnitude smaller than the measured k value at this temperature. The above indicates that the k(T) dependence is strongly influenced by the phonon contribution, as described elsewhere [20]. The large phonon contribution to k(T) is present as well in REF, due to the very low density of this sample, even when the electrical resistivity values are lower. Therefore, grain boundaries seem to be a more plausible source of electron scattering for the decrease in k(T), and such a behavior is much less marked in all SPS samples with high density. It is even less marked in sample S750, which has a smaller width of deoxygenated shell due to the decreasing of TD . As it was previously mentioned, grain boundaries play an important role in limiting the transport properties of these materials. Thus, the features occurring at the grain-boundaries have enormous influence on the behavior of magnetic field dependence of the critical current density, Jc (Ba). From several I − V curves, taken at 77 K and under different applied magnetic fields, we were able to build a normalized Jc (Ba)/Jc (0) versus applied magnetic field diagram, as shown in Figure 7. The measurements were performed at the same reduced temperature t = T /To f f = 0.94 for comparison reasons. We first mention here that the ratio Jc (Ba)/Jc (0) abruptly decreases with increasing applied magnetic field Ba , and displays a clear Josephson-like behavior. We also mention that values of Jc (Ba)/Jc (0) in the SPS samples S750 and S830 are smaller than those measured in the reference sample, indicating the role played by the deoxygenation in the transport properties of the SPS samples. The estimated values for the superconducting critical current density at zero applied magnetic field, Jc (0), was found to be ∼ 1.4, 2.4, and 1.3 A/cm2 for REF, S750, and S830 respec-



237

tively. This indicates that the high density in SPS samples had a major role in improving the connectivity between grains, except in the sample consolidated at a higher temperature, TD = 830 ◦ C, which has a higher width of deoxygenated shell. It is important to note that the decrease in Jc (Ba)/Jc (0) is more pronounced in sample S830 at the same range of applied magnetic field, indicating that the flux penetration through grain boundaries is significant when the consolidation temperature TD increase. The flux penetration through grain boundaries was investigated from a particular value of the applied magnetic field, B0 , which is related with the value where Jc (Ba)/Jc (0) decreases ∼ 50%. Such value is taken as the critical field of the Josephson junctions, B0 = Φ0 /2λL Lg , where Φ0 is the flux quantum, λL is the London penetration depth, and Lg is the grain size. Values of B0 are very small in all samples (see Table 1). If we assume that the SPS process unaffects significantly the size of the grains, then the differences on B0 can be related to changes in the λL of the materials. Therefore, an increase of B0 leads to a decrease of λL . This behavior can be related to changes in the width of the oxygen deficiency shell near the grain boundaries, as suggested previously.

Figure 5. Temperature dependence of the electrical resistivity of reference sample and the all SPS sample in the vicinity of the superconducting transition. The insets display the ρ(T) curves in all range of temperature. In any event, the analysis of the above results indicates that increasing the consolidation temperature TD results in a progressive deoxygenation of the SPS samples, and that is much more pronounced on the grain-boundary. A way to partially restore partially the oxygen content is to subject the SPS samples to an additional post-annealing heat treatment (PAHT) in air. This point is discussed below.

4.3.

The Effect of Post-Annealing Heat Treatment (PAHT)

For the sake of brevity, we only refer to the values of time and temperature that showed the best results during the post-annealing heat treatment. Thus, the SPS samples were subjected to a PAHT performed in air at 750 ◦ C for 5 min. For this, the specimens were cut from SPS


238


Figure 6. Temperature dependence of the thermal conductivity of the reference sample REF and the SPS samples S750, and S830.

Figure 7. Normalized Jc(Ba)/Jc (0) curves measured in samples REF, S750, and S830. samples S830, S800, and S750, which are referred to as P830, P800, and P750 respectively. All analysis were performed by using the same characterization techniques employed in the previous section. Firstly, we can ask ourselves if the post-annealing heat treatment influences the microstructure of our samples. The X-ray diffraction patterns taken on powders of all postannealed samples are shown in Figure 8. The analysis indicates a single-phase material with nominal composition Bi-2223. Besides, no extra reflections belonging to the infinite layer compound Ca1−x Srx CuO2 were observed. We can also mention that the post-annealing heat treatment did not alter the chemical composition of the SPS samples and single-phase



239

Figure 8. The X-ray diffraction patterns taken on powders of REF, S750 and the postannealed sample P750. specimens were obtained even when the post-annealing sintering was performed for just 5 min. However, the improvements on superconducting properties in all SPS samples was observed after the PAHT. In order to address the influence of PAHT on the magnetic properties, the χ(H) dependence of the SPS samples before (S750, S830) and after (P750, P830) the PAHT are displayed in Figure 9. After the heat treatment, all SPS samples presented a higher magnetic signal at the selected temperature, and the values of the first intergranular critical field increased until Hc1 j ∼ 3 and 4 mT for P750 and P830 respectively. These results clearly demonstrate an improvement in intergranular properties of the samples SPS. Such improvements are associated with the decrease in the width of the deoxygenation shell and the increase in the superconducting volume fraction of the grains. Furthermore, we believe that the reoxygenation of the SPS samples is a process that is accompanied by the establishment of conductive filamentary paths along the grain boundaries with the consequent contribution to facilitate the flow of shielding currents at intergranular level. It is reasonable to point out that the effect of the reoxygenation process and the conductive filamentary paths formation must be reflected in the transport properties of these samples. The temperature dependence of the electrical resistivity ρ(T) of the post-annealed samples is shown in Figure 10. The first point is its T -linear, metallic-like behavior on the normal region for all samples. It is also important to note that the onset superconducting critical temperature, Ton , was found to take the same value ∼ 114 K for all samples as well as To f f ∼ 103 K. In addition, after the heat treatment all, samples exhibits ρ(300K) values quite smaller than those corresponding to the original samples (4.3, 5.3, and 6.8 mΩcm for P750, P800, and P830 respectively), although slightly smaller than the one of the reference sample (∼ 9 mΩcm). These combined results suggest that the reoxygenation process has restored many of the intragranular and intergranular properties of the SPS Bi-2223 samples. On the other hand, a preliminary analysis from the electronic contribution to the thermal conductivity indicates that by increasing B0 , there is a decrease in λL in the post-annealed


240


samples, which is result is related to changes in the width of the oxygen-deficient shell of the superconducting grains. However, the most important result is related to the values of critical current density at 77 K. It is very important to mention here that Jc of our SPS samples has increased significantly after the PAHT. The Jc value in sample S750 was increased from ∼ 10 A/cm2 to ∼ 128 A/cm2 after the post-annealing treatment (sample P750), a value six times higher than the reference sample (REF). Besides, the PAHT resulted in an even higher increase of Jc in sample S830, from ∼ 2.1 A/cm2 to ∼ 58.1 A/cm2 . These results indicate a ∼ 10 to 25fold increase of Jc in our SPS samples. Everything seems to indicate that, in these samples, a delicate balance among reoxygenation and changes in their microstructures determines the effectiveness of the pinning centers at the grain boundaries and consequently high values of Jc .

Figure 9. Magnetic field dependence of the dc magnetic susceptibility measured at T = 2K in pellets of samples S750, S830, P750 and P830. See text for further details.

4.4.

The Effect of the Plungers/Die Material

Usually, the spark plasma sintering process is conducted by using a setup built with highdensity graphite die and plungers, denominated here as all-graphite setup (A-GS) [21]. The A-GS setup has some disadvantages associated with its thermal and mechanical properties [21]. For instance, barium ferrite materials obtained by using the A-GS set up exhibit carbon contamination and reduction of Fe3+ to Fe2+ , due to the diffusion of carbon monoxide within the samples [22]. More recently, Mackie et al. have demonstrated the carbon uptake by samples of Sm(Co,Fe,Cu,Zr)z obtained by the SPS method [23], as inferred from carbon distribution maps of the samples obtained by using Electron Probe Micro-Analysis (EPMA). The results indicate a carbon contamination within the samples, a feature especially pronounced at the surface of the materials. Such undesired results may be minimized by replacing graphite by other materials. As far as this point is concerned, the use of steel has been proposed in order to reduce the electrothermal loss during the SPS process, a



241

Figure 10. Temperature dependence of the electrical resistivity ρ(T) of post-annealed samples P750, P800, and P830. mechanism known as reduced electrothermal loss SPS (RETL-SPS) [21]. Samples were then consolidated by using two setups comprised of different materials: all-steel and allgraphite. Finite element simulations (FEM) were performed to provide extra information regarding the distribution of temperature within the samples, as well as other complementary characterizations. The temperature profiles of samples S750, H700, and H750 are shown in Figure 11 (a). In all cases, the temperature at the position of the thermocouple increases as the processing time is increased. For samples H700 and H750 the temperature increases non-linearly during the heating ramp, t ≤ tr , in contrast to the linear behavior observed in the profile of sample S750. During the consolidation time, tr ≤ t ≤ tD (see the insets of Figure 11 (a)), there is a small, but perceptible difference between the experimental temperature profiles. In sample H700, when the sintering time attains tD , the PID temperature controller of the SPS machine hardly stabilizes the system, i.e., the temperature oscillates between ± 25 ◦ C in the vicinity of TD = 725 ◦ C. This behavior may be ascribed to a combination of physical properties of the AISI H13 steel, e.g., electrical resistivity and thermal conductivity, as discussed below. A possible way to minimize the overheating observed in the temperature profile of sample H700 is decreasing the heating rate or, equivalently, to increase the heating time. These changes were made, resulting in sample H750, and its temperature profile is displayed in Figure 11 (a). For this particular case, the obtained results are of interest. We first mention that the temperature measured at the position of the thermocouple increases nonlinearly, reaching a plateau at T ∼ 725 ◦ C for t ≥ 8 min. At t = tr = 15 min, the temperature starts to oscillate in a range of ± 25 ◦ C close to TD = 750 ◦ C. On the other hand, when the allgraphite setup is used, the consolidation the sample S750 occurs according to the predefined conditions. As observed in Figure 11 (a), the temperature increases linearly and reaches a maximum of 775 ◦ C during the time interval of the consolidation process. In this case, the PID controller stabilized well and the consolidation temperature of TD = 750 ◦ C varied little, between 7 ≤ t ≤ tD.


242


Figure 11. (a) Temperature profile during the SPS consolidation of samples S750, H700, and H750; (b) X-ray diffraction patterns taken from the powders obtained after crushing the SPS pellets H700 and H750; (c) Temperature dependence of the electrical resistivity of post-annealed SPS samples H700 and H750. (see text for details).



243

The next step is to inspect the influence of the above sintering conditions on the phase content of the consolidated samples. We first mention that the optimum sintering temperature for producing nearly single-phase Bi-2223 is Ts ∼ 845 ◦ C [24]. Also, samples obtained at temperatures below or above Ts usually exhibit extra phases as Ca2 PbO4 , Bi-2201, and Bi-2212 [25,26]. Accordingly, figure 11 (b) shows the X-ray diffraction patterns taken from powders the samples H700 and H750. The results indicate that most of the intense peaks in the pattern of sample H750 belong to the Bi-2212 phase. At first approximation and by considering that all samples are comprised of only two phases (Bi-2223 and Bi-2212), then a roughly estimate of the volume fraction of the Bi2 Sr2 CaCu2 O8+δ (Bi-2212) phase yields ∼ 6 and 60% in samples H700 and H750, respectively. The presence of traces of the Bi-2212 phase in the X-ray pattern of consolidated Bi1.65 Pb0.35 Sr2 Ca2 Cu3 O10+δ samples by SPS is an expected result, as reported elsewhere [19, 27]. However, the above results indicate that the volume fraction of the Bi-2223 phase in sample H750 has been altered significantly after the SPS consolidation. Such a large amount, close to 60%, of the Bi-2212 phase indicates that appreciable volume fraction of the starting Bi-2223 phase has been converted during the SPS process. On the other hand, it is expected that all the above results have their counterpart on the transport properties of the obtained samples. Following this discussion, the temperature dependence of the resistivity, ρ(T), measured in pellets of samples H700 and H750 after the post-annealing heat treatment, are displayed in Figure 11 (c). In addition, the ρ(T ) dependence of sample P750 (see Table 1) was also included for comparison reasons. As expected, the behavior of ρ(T ) dependence in sample H750 in the transition region clearly indicates the presence of two superconducting transitions. The first, close to 110 K and related to the high-Tc phase 2223, and the other one near 80 K regarding the low-Tc phase 2212. These results are in line with the X-ray analysis as discussed above. Another important result displayed in the Figure 11(c) is related to the behavior of the ρ(T ) dependence in the normal-state region for T ≥ 110 K. It was found that the value of ρ(300K) of sample H700 is two times greater than in sample P750. Notice that in these samples the phase composition is almost the same. Thus, the discrepancies observed between values of ρ(300K) can be related to differences in the volume density, ρv , in these samples (see the Table 1). In this case, sample H700 exhibits a volume density of 4.8 g/cm3 , which assures a low grain connectivity than in sample P750 where ρv = 5.5 g/cm3 .

Final Remarks Pre-reacted powders of (Bi-Pb)2 Sr2 Ca2 Cu3 O10+δ (Bi-2223) were consolidated by using the SPS technique under vacuum and at different consolidation conditions. The results of X-ray powder diffraction indicated that the materials consist of the Bi-2223 dominant phase. The density of the SPS pellets reached ∼ 85% of the theoretical value, even when the compacting pressure was only 50 MPa. We have also found that both the normal and superconducting transport properties of the SPS samples are strongly dependent on their oxygen content and microstructures, a feature much more marked in those consolidated at 800 and 845 ◦ C. The combined results point out to the occurrence of grains with core-shell morphology, where the shell is oxygen deficient. We have argued that the width of the shell increases with increasing consolidation temperature, resulting in samples with very low Jc values at


244


T = 77 K. Post-annealing heat treatment (PAHT) performed in air and at 750 ◦ C for a brief time interval (5 min), gives rise to single phase materials with Jc (77K) ∼ 130 A/cm2 , a value six times higher than the sample obtained by the traditional ceramic method. The results discussed here strongly indicated that a PATH is an important step towards increasing the transport critical current across grain boundaries in SPS samples for future commercial applications, e.g., resistive current-fails devices and superconductor magnets. On the other hand, the influence of plungers/die built with different materials on the superconducting properties of Bi-2223 ceramic was investigated. The experimental temperature profiles obtained during the SPS consolidation were successfully fitted to the generated curves from a finite element simulation model integrated with a PID control. The comparison of the experimental and theoretical curves demonstrated the good accuracy of the proposed model. Also, the samples consolidated by using the all-graphite setup display higher superconducting volume fraction of the Bi-2223 phase when compared to those consolidated with the all-steel setup, and the carbon of the mold has no influence on promoting the oxygen deficiency of the shell layer. Even though SPS samples obtained by using stainless steel dies were found to have lower superconducting volume fraction, the use of this setup raises the possibility of achieving higher values of uniaxial compacting pressure during the SPS process. Finally, all the above results strongly indicate that spark-plasma sintering is a promising technique to obtain Bi-2223 ceramic samples with well-controlled microstructure. These samples can be considered for power applications as resistive current-fault limiter. However, the core-shell morphology of the grains is the main limitation to reach this goal. Thus, further strategies are being tested in order to avoid this undesired result.

References [1] Tan T. T., Li S., Cooper H., Gao W., Liu H. K., Dou S. X. 2001. "Characteristics of micro-texture and meso-texture in (Bi, Pb)2 Sr2 Ca2 Cu3 O1 0 superconducting tapes" Supercond. Sci. Technol. 14:471-478. [2] Hilgenkamp H., and Mannhart J. 2002. "Grain boundaries in high-T c superconductors" Rev. Mod. Phys. 74:485-549. [3] Jill Glass S., and Ewsuk Kevin G. 1997. "Ceramic powder compaction" MRS Bulletin:24-28. [4] Govea-Alcaide E., Jardim R. F., and Muné P. 2005. "Correlation between normal and superconducting transport properties of Bi1.65 Pb0.35 Sr2 Ca2 Cu3 O10+δ ceramic samples." Physica C 423:152-162. [5] Carneim Robert D., and Messing Gary L. 2001. "Response of granular powders to uniaxial loading and unloading." Powder Technology 115:131-138. [6] Chikui N., Furuhata H., Yamaguchi N., and Ohashi O. 2005. "Guidelines for using pulsed electric current bonding on stainless steel" J. Jpn. Metals 69:715-718.



245

[7] Machado I. F., Girardini L., Lonardelli I., and Molinari A. 2009. "The study of ternary carbides formation during SPS consolidation process in the WC−Co−steel system" Int. J. Refract. Met. Hard Mater. 27:883-891. [8] Chen X. J., Khor K. A., Chan S. H., and Yu L. G. 2004. "Overcoming the effect of contaminant in solid oxide fuel cell (SOFC) electrolyte: spark plasma sintering (SPS) of 0.5 wt.% silica-doped yttria-stabilized zirconia (YSZ)" Mater. Sci. Eng. A 374:6471. [9] Muñoz S., and Anselmi-Tamburini U. 2010. "Temperature and stress fields evolution during spark plasma sintering processes." J. Mater. Sci. 45:6528-6539. [10] Anselmi-Tamburini U., Gennari S., Garay J. E., and Munir Z. A. 2005 "Fundamental investigations on the spark plasma sintering/synthesis process: II. Modeling of current and temperature distributions." Mater. Sci. Eng. A 394:139-148. [11] Wang C., Cheng L., and Zhao Z. 2010. "FEM analysis of the temperature and stress distribution in spark plasma sintering: Modelling and experimental validation." Comp. Mater. Sci. 49:351-362. [12] Ziegler J., and Nichols N. 1942. "Optimum settings for automatic controllers, Trans." ASME 64:759-765. [13] Giannini E., Garnier V., Gladyshevskii R., and Flukiger R. 2004. "Growth, structure, and superconducting properties of Bi2 Sr2 Ca2 Cu3 O1 0 and (Bi,Pb)2 Sr2 Ca2 Cu3 O10−y crystals" Supercond. Sci. Technol. 17:220-226. [14] MacManus Driscoll J.L., Pin Chin Wang, Bravman J. C., and Beyers R. B. 1994. "Phase equilibria and melt processing of Bi2 Sr2 Ca1 Cu2 O8+x tapes at reduced oxygen partial pressures" Appl. Phys. Lett. 65:2872-2874. [15] Rao C. N. R., Ganapathi L., Vijayaraghavan R., Rao G. R., Murthy K., and Moha n Ram R. A. 1998. "Superconductivity in the Bi2 (Ca,Sr)n+1 Cun O2n+4 (n=1, 2, or 3) Series: Synthesis, Characterization and Mechanism." Physica C 156:827-833. [16] Fujii T., Watanabe T., and Matsuda A. 2001. "Comparative study of transport properties of Bi2 Sr2 Ca2 Cu3 O10+δ and Bi2 Sr2 CaCu2 O8+δ single crystals" Physica C 357:173-176. [17] Paturi P., Raittila J., Huhtinen H., Huhtala V. P., and Laiho R. 2003. "Size-dependent properties of YBa2 Cu3 O6+x nanopowder" J. Phys. Condens. Matter 15:2103-2114. [18] Pérez F, Obradors X, Fontcuberta J, Bozec X, and Fert A. 1996. "Magnetic flux penetration and creep in a (Y,Sm)Ba2 Cu3 O7 ceramic superconductor" Supercond. Sci. Technol. 9:161-175. [19] Govea-Alcaide E., Machado I. F., Bertolete-Carneiro M., Muné P., and Jardim R. F. 2012. "Consolidation of Bi-2223 superconducting powders by spark plasma." J. Appl. Phys. 112:113906-113914.


246


[20] Castellazi S., Cimberle M. R., Ferdeghini C., Giannini E., Grasso G., Marre D., Putti M., and Siri A. S. 1997. "Thermal conductivity of a BSCCO (2223) c-oriented tape: a discussion on the origin of the peak" Physica C 273:314-322. [21] Chennoufi N., Majkic G., Chen Y. C., and Salama K. 2009. "Temperature, Current, and Heat Loss Distributions in Reduced Electrothermal Loss Spark Plasma Sintering." Met. Mater. Trans. A 40A:2401-2409. [22] Ovtar S., Gallet S., Minier L., Millot N., and Lisjak D. 2014. "Control of barium ferrite decomposition during spark plasma sintering: Towards nanostructured samples with anisotropic magnetic properties." J. Eur. Ceram. Soc. 34:337-346. [23] Mackie A. J., Hatton G. D., Hamilton H. G. C., Dean J. S., and Goodall R. 2016. "Carbon uptake and distribution in Spark Plasma Sintering (SPS) processed Sm(Co, Fe, Cu, Zr)z ." Mater. Lett. 171:14-17. [24] Muné P., Govea-Alcaide E., and Jardim R. F. 2003 "Influence of the compacting pressure on the dependence of the critical current with magnetic field in polycrystalline (Bi-Pb)2 Sr2 Ca2 Cu3 Ox superconductors." Physica C 384:491-500. [25] Pandey D., Singh A. K., Srivastava P. K., Singh A. P., Inbanathan S. S. R., and Singh G. 1997. "Towards the rapid synthesis of pure 2223 powders with Bi2x Pbx Sr2 Ca2 Cu3 0y compositions by semi-wet methods II. Optimization of Pb content using Pb-Sr-Ca carbonate precursors." Physica C 241:279-291. [26] Majewski P. 1997. "Phase diagram studies in the system Bi–Pb–Sr–Ca–Cu–O–Ag." Supercond. Sci. Technol. 10:453–467. [27] Badica P., Aldica G., Groza J. R., Bunescu M.-C., and Mandache S. 2002. "Reactivefield-assisted-sintering of freeze-dried powders in the BSCCO system, Supercond. Sci. Technol. 15:32-42.




Chapter 11

FLUCTUATION INDUCED EXCESS CONDUCTIVITY OF Bi2Sr2CaYXCu2Oy SUPERCONDUCTORS A. Sedky Physics Department, Faculty of Science, Assiut University, Assiut, Egypt

ABSTRACT We report here the fluctuation induced excess conductivity of Bi2Sr2CaYxCu2Oy with various x values (0.00 < x < 0.50). This work is done by using the reported data of Sedky, J Superconductor and Novel Magnetism 29, 1475 (2016), and with the help of Anderson and Zou relation. The mean field temperature Tcmf, deduced from the peak of dρ/dT versus temperature plot, is increased by Y addition up to 0.15 followed by a decrease with further increase of Y up to 0.50 as well as crossover temperature T o. The logarithmic plots of ∆σ and reduced temperature Є reveal two different values of the order parameter exponents corresponding to crossover temperature. The first exponent is obtained in the normal field region at a temperature of (Tcmf < T < 2 Tcmf)), while the second exponent is obtained in the mean field region at a temperature of (T > Tcmf). The order parameter dimensionality OPD is shifted from (3D) to (2D) for pure sample as the temperature reduced towards the critical temperature Tc, but it is shifted from (1D/qusi-1D/2D) to (qusi-1D/qusi-2D) for Y doped samples. The interlayer coupling, coherence lengths and anisotropy are calculated and their values are decreased by Y addition up to 0.30 followed by an increase with further increase of Y up to 0.50. The vice is versa for critical fields and critical current density against Y addition. It is believed that some of secondary phases rich by Cu may be formed by Y addition and leads to point defects in the order of nanometers. These defects act as a vortex pinning centers which improve the critical values of the field and current.



Corresponding E-mail: [email protected].


248

A. Sedky

Keywords: excess conductivity, fluctuation, order parameter, critical fields and critical current

1. INTRODUCTION The discovery of Bi2Sr2CaCu2Oy (Bi: 2212) superconducting system with critical temperature TC of 87 K has generated a great scientific interest among researchers. It is accepted that Bi: 2212 superconductors and related systems are used in fabrication of wires and tapes due to their higher values of critical current density and critical fields. The substitution of rare-earth elements R in place of Ca in Bi: 2212 superconductors lead to structural stability and normally used to improve their superconducting properties [1]. It is also helps for understanding the type of charge carriers as well as carrier concentration [2-7]. However, some of studies show an improvements in these properties by the low amount of not exceeds 0.20 of rare-earth elements [8, 9-15]. Furthermore, an improvement of critical temperature and critical current density of Bi1.6Pb0.4Ca1.1Cu2.1RxO8+δ superconducting system is reported (R = Pr, Yb, Sm, Nd and La) [16-20]. The high temperature superconductors in particular exhibit anisotropy and small coherence length together with elevated values of critical temperatures Tc. These materials have a great effects on the fluctuations of superconducting order parameter. These effects have been early observed in the conductivity versus temperature curves as induced excess conductivity. Therefore, the fluctuation induced excess conductivity is an important parameter for exploring the normal state properties above T c in these materials [21-24]. The fluctuation induced conductivity (FIC) analyses reveal that the contribution of excess conductivity is due to Gaussian fluctuation in the mean field region as well as the critical field region [25]. Gaussian fluctuation is probably dominant in the temperature region above the mean field temperature Tcmf when the fluctuation in the order parameter is small and the interactions between Cooper pairs can be neglected. While the critical fluctuation occurs in the critical field region below the Tcmf when the fluctuation in the order parameter is large and the interactions between Cooper pairs is considered [26]. The dimensional exponents are found to be zero dimensional (0D), one dimensional (1D), two dimensional (2D) and three dimensional (3D) [27-28]. It seems that the dimensional crossover takes place between any two different regions, and it is mainly obtained above Tc at a crossover temperature To. However, the OPD of BSCCO samples are 2D dimensional and the crossover is usually occurs either from 3D to 2D or from 1D to 2D [29-33, 34]. Recently A. Sedky et al. have investigated the structural and superconducting properties of Bi2Sr2CaYxCu2Oy superconductors with various x values (0.00 ≤ x ≤ 0.50)


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 249 by using the resistivity and ac susceptibility measurements [35]. It is found that addition of Y3+ does not influence the phase purity of Bi: 2212, but the c- parameter and orthorhombic distortion are affected. Furthermore, the critical temperatures Tc (R = 0) are increased from 92 K for Y = 0 up to 106 K by 0.15 of Y, followed by a decrease with increasing Y up to 0.50. It is also noted that the resistivity of the samples with Y = 0.075 and 0.15 reaching to zero resistivity (ρ = 0) only at low dc currents up to 2 mA. Interestingly, the onset temperature of diamagnetism are kept at 94 K by increasing Y up to 0.30, followed by a decrease to 86 K at Y = 0.50. Moreover, the superconducting volume fraction and critical currents are improved by Y addition up to 0.15, followed by a decrease with increasing Y up to 0.50. These results are explained in terms of the effects of excess oxygen, secondary phases and carrier concentration per Cu ions which are produced by Y addition in Bi: 2212 system. As a continuation of the above work, we reported here the fluctuation induced conductivity produced by Y addition on the same batch of samples. We have restricted our analysis to the mean field regime and crossover behavior, and tried to extract some physical parameters such as coherence lengths, anisotropy, interlayer coupling, dimensional exponent, critical magnetic field and critical current density.

2. EXPERIMENTAL DETAILS The ingredients of the materials Bi2O3, SrO,Y2O3, CaCO3 and CuO of 4N purity (Bi2Sr2CaYxCu2Oy) are thoroughly mixed in required proportions and calcined at 820 oC in air for 16 h. This exercise is repeated three times with intermediate grinding at each stage. The resulting powder is reground, mixed, pressed into pellets under a load of 6 tons. After that the pellets are then sintered in air at 840 oC for 50 h and left in the furnace for slow cooling to room temperature. The maximum rate of heating/cooling is nearly about (5o/mints). The phase purity of the samples and structural morphology are examined by using X-ray diffraction (XRD) and scanning electron microscope (SEM). The electrical resistivity of the samples is obtained using the standard four-probe technique in closed cycle refrigerator [cryomech compressor package with cryostat Model 810-1812212, USA] within the range of (10-300 K). Nanovoltameter Keithley 2182, current source Keithley 6220 and temperature controller 9700 (0.001 K resolution) are used for resistivity analysis.

3.1. Theoretical Background The excess conductivity ∆σ due to thermal fluctuation is defined by the deviation of the measured conductivity of σm (T) from the normal conductivity σn (T) as follows;


A. Sedky

250

  (

1

m



1

n

)  m n

(1)

where ρm and ρn are the measured and normal resistivity. ρn is obtained from the measured resistivity ρm at T ≥ 2Tc by applying the least square method to the Anderson and Zou relation,

 n (T )  A  BT

[36]. In order to estimate the paraconductivity,

Aslamazov and Larkin (AL) deduced the following relation for the fluctuation induced excess conductivity ∆σ [37] as; ∆σ = A ε -λ

(2) 2

e e2 Here, A  and λ = 0.5 for 3D conductivity, and A  and λ = 1.0 16d 32 c (0) for 2D conductivity, e is the electronic charge, d is the interlayer spacing between two successive CuO2 planes, ћ is the reduced Planks, constant, ξc (0) is the c-axis 3D coherence length at zero temperature and Є is the reduced temperature given by [38-40];

T  Tcmf ε= Tcmf

(3)

We have followed the dρ/dT versus T plot to obtain the values of Tcmf in terms of the peaks of dρ/dT. For polycrystalline samples, the modified equations for 2D and 3D fluctuations are expressed as [41]; 1

 3 D 

 e2  2 32 p (0)

 2 D 

1  e2 8c4 (0) 1 2 1  [ 1  ( 1   )  2 4 16d d 2  ab (0)

(4a)

1



(4b)

where ξab(0) is the coherence length at 0 K across the ab- plane and ξp(0) is the effective coherence length at 0 K. On the other hand, the crossover behavior from 2D-3D occurs at a temperature T0 given by [40];


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 251 T0  Tcmf exp(

2 c (0) 2 ) d

(5)

where ξc (0) is given by [42- 43];

(6) where J is the interlayer coupling, and expressed by [44,45];

J  ln(

T0 ) 2Tcmf

(7)

4. RESULTS AND DISCUSSION The resistivity versus temperature curves of the considered samples are given in Figure 1 (a-e). The normal resistivity is found to be linear as the temperature reduced from room temperature down to a certain temperature T B ~ 2Tcmf. In this region follows the above formula,

 n (T )  A  BT .

n (T )

TB ~ 2Tcmf, and it is defined as the

temperature below which the Cooper-pair formation starts [24, 44]. As the temperature is further reduced beyond the normal state region, the rate of resistivity drop becomes entirely different as compared to this region. This is due to increasing Cooper pair formation, and consequently the fluctuation induced conductivity in this region follows the Aslamazov and Larkin (A-L model) [38]. ρn(T) is calculated by using the values of A and B parameters obtained from the fitting and shown in Figure 1. We have extrapolated the linear fit of the normal state resistivity to the lower temperature regime as shown by straight lines in Figure 1. The straight columns drawn in the curves indicate the width of temperatures fitting. One of them occurs at a temperature close to TB and the second occurs at a temperature close to room temperature. The mean field temperatures Tcmf for all samples are estimated from the peak of dρ/dT against temperature plot shown in Figure 2. Similar values are listed in Table 1.


252

A. Sedky

Figure 1(a). Resistivity versus temperature for Bi2Sr2Ca YxCu2Oy samples(x = 0.00).

Figure 1(b). Resistivity versus temperature for Bi2Sr2Ca YxCu2Oy samples (x = 0.075).

Figure 1(c). Resistivity versus temperature for Bi2Sr2Ca YxCu2Oy samples(x = 0.15).


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 253

Figure 1(d). Resistivity versus temperature for Bi2Sr2Ca YxCu2Oy samples (x = 0.30).

Figure 1(e). Resistivity versus temperature for Bi2Sr2Ca YxCu2Oy samples (x = 0.50).

Table 1. OD, Tc, Tcmf, T0, ξc(0), ξp, ξab, γ, λI and λII for Bi2Sr2CaYxCu2Oy samples x 0.00 0.075 0.15 0.30 0.50

(b-a)/b OD 0.0020 0.0015 0.0013 0.0009 0.0004

Tc (K) 92 99 106 86 56

Tcmf (K) 96 102 114 95 74

To (K) 104 112 125 105 82

J 0.61 0.60 0.60 0.59 0.59

ξc(0) (Å) 6.02 5.97 5.96 5.89 5.88

λI

λII

0.57 3D 1.19 qusi-1D 1.05 2D 1.01 2D 1.18 qusi-1D

1.08 2D 1.14 qusi-1D 1.22 qusi-1D 1.24 qusi-1D 0.69 qusi-2D

By using the values of ∆σ and reduced temperatures ε, we have plotted ln ∆σ against ln ε for all samples, see Figure 3. It is evident from the fitting that there is one distinct


254

A. Sedky

change in the slope of each plot. The corresponding temperature where the slope change occurs is designated as the crossover temperature T o. Therefore, the crossover temperature along with two different exponents are obtained from each plot with an accuracy of + 1 K. Anyhow, the different values of T c, Tcmf and To against Y content are shown in Figure 3 (a). Similar values are listed in Table 1. It is evident from the Figure that both Tc, Tcmf and To are increased by increasing Y content up to 0.30, followed by a decrease with further increase of Y up to 0.50.

Figure 2. dρ/dT versus temperature for Bi2Sr2Ca YxCu2Oy samples.

Figure 3. Ln ∆σ against Ln ε for Bi2Sr2Ca YxCu2Oy samples.


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 255 However, the first exponents are obtained above the mean field region at a temperature range of (-1.02 ≥ ln ε ≥ - 2.44) for x = 0.00, (- 0.95 ≥ ln ε ≥ - 2.36) for x = 0.075, (-1.00 ≥ ln ε ≥ - 2.12) for x = 0.15, (0.50 ≥ ln ε ≥ - 2) for x = 0.30, (- 0.94 ≥ ln ε ≥ - 1.97) for x = 0.50. The values of OPD are 0.57 (3D) for the pure sample, and shifted to (qusi-1D/2D) for all Y doped samples. The OPD are 1.19 (qusi-1D) for x = 0.075, 1.05 (2D) for x = 0.15, 1.01 (2D) for x = 0. 30 and 1.18 (qusi-1D) for x = 0.50. While, the second exponents are obtained in the mean field region above T c at a temperature range of (-2.73 ≥ ln ε ≥ - 3.82) for x = 0.00, (- 2.77 ≥ ln ε ≥ - 4.72) for x = 0.075, (- 0.2.3 ≥ ln ε ≥ - 3.91) for x = 0.15, (-2.48 ≥ ln ε ≥ - 3.87) for x = 0.30, (- 2.19 ≥ ln ε ≥ - 3.58) for x = 0.50. The values of OPD are (2D) for pure sample, and shifted to (qusi-1D/qusi-2D) for all Y doped samples. The OPD are 1.08 (2D) for x = 0.00, 1.14 (qusi-1D) for x = 0.075, 1.22 (qusi-1D) for x = 0.15, 1.24 (qusi-1D) for x = 0. 30 and 0.69 (qusi-2D) for x = 0.50. Figure 3(b) shows the variation of the order parameter as a function of Y content, and similar values are listed in Table 1. These results indicate that the crossover occurred nearly from (qusi-1D/2D) to (qusi-1D/ qusi-2D) for all Y doped samples except x = 0.00 sample, in which the crossover occurred from 3D to 2D. To our knowledge the present analysis of fluctuation induced conductivity may be reported for the first time. Normally, the OPD of pure BSCCO systems are 2D in the critical field region. It has been reported that the crossover occurred either from 3D to 2D or from 2D to 1D in these systems due to the effect of radiation [26, 34, 45, 46]. Actually, the critical field region is controlled by the critical fluctuation results from the small mean free path of the charge carriers and also the short coherence length produced by changing carrier concentration [26, 47]. This is of course, due to the formation of high Tc phase produced by Y addition up to 0.15 as reported for Tc and hardness variations [35]. It is also reported that the carrier concentration /Cu ions is decreased by Y substitution up to 0.30, followed by a decrease at Y = 0.50 [35], in consistent with the present behavior for OPD. On the other hand, the interlayer coupling strength k is calculated by using equation (7), and their values are used for calculating c-axis coherence length at 0 K ξc (0), d = c/2 = 18 Ǻ for Bi: 2212 systems [48]. Also, for polycrystalline samples, ξp(0) given by [40]; 1

1 1 1 1 8  [  ( 2  2 )2 ]  p (0) 4 c (0) c (0)  ab (0)

(8)

ξab (0) is calculated in terms of ξp(0) and ξc(0) values. Then, anisotropy parameter



 ab (0) , are easily obtained. However, the variation of J and  against Y content is c (0)

shown in Figure 4 (b), and similar values are listed in Tables 1 and 2.


A. Sedky

256

Table 2. Bc, Bab and J (0K) for Bi2Sr2Ca YxCu2Oy samples x 0.00 0.075 0.15 0.30 0.50

ξp (Å) 5.41 5.36 5.25 5.11 5.48

ξab (Å) 5.06 4.89 4.77 4.61 5.11

γ 0.84 0.82 0.80 0.78 0.87

Bc (T) 1287.4 1378.5 1448.7 1551 1262.3

Bab (T) 1093 1134.8 1165.3 1207.8 1080.5

Jc (0K) (A/cm2) 1.59E  104 1.61  104 1.64  104 1.68  104 1.57  104

Figure 4(a). Tc, Tcmf and To versus Y content for Bi2Sr2Ca YxCu2Oy samples.

Figure 4(b). Order parameter, interlayer and anisotropy versus Y for Bi2Sr2Ca YxCu2Oy samples.


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 257 The values of  are decreased by increasing Y content up to 0.30, followed by a sharp increase at Y = 0.50. It is decreased from 0.85 for Y = 0.00 to 0.78 for Y = 0.30, followed by an increase up to 0.90 for Y = 0.50. While J is improved by increasing Y up to 0.30, followed by a decrease at Y = 0.50. Figure 4 (c) shows the behaviors of ξc(0), ξab(0) and ξp(0) parameters against Y content, and similar values are listed in Tables 1 and 2. It is clear that ξc(0) is gradually decreased by Y content up to 0.50, in agreement with the behaviors of c-axis against Y content [35]. While ξab(0) and ξp(0) are decreased by Y content up to 0.30, followed by an increase at Y = 0.50, in agreement with the behaviors of Tc, carrier density, OPD behavior and  against Y content. Actually, the Bi: 2212 system is essentially 2D with two Cu-O2 planes which are manifest in the lower values of J and higher degree of anisotropy as compared to Y: 123 systems [48]. However, the increase of J against Y addition suggested that the system has lower anisotropy, and consequently the carrier density should be increased toward the critical value. It has been also reported that the doping up to considerable level produces depletion for the excess of oxygen, thereby improving the metalcity of Bi-O layer [37]. The upper critical fields along the c-axis and a-b plane, and critical current density Jc (0 K) is estimated by the following relations [23, 49-50];

BI I (c) 

0 0 , BI I (ab)  2 2 c (0) ab (0) 2 ab (0)

J c (0) 

20 6 (0) p (0)

(10)

2

where 0 is quantum flux given by 0 

(9)

h 2e

 2.07  1015 ( web / m 2 ) , and µ is London

penetration depth at 0 K which is about 250 nm for Bi:2212 superconductors [51]. However, the behaviors of B and J(0K) against Y content are shown in Figures 5(a) and 5 (b). Similar values are listed in Table 2. It is clear that Bab, Bc and Jc are increased by Y content up to 0.30, followed by a decrease for Y = 0.50. This is due to the enhancement of flux pinning ascribed by increasing the pinning centers in these types of samples [51, 52]. Anyhow, a universal dome- shaped Tc versus carrier concentration has been observed in Bi:2212 system [52]. It is has been also reported that Tc increases with increasing carrier concentration, passes through a maximum, decreases and becomes zero above an optimum value of concentration [53]. Therefore, the reason for the enhancement in B and J c up to Y = 0.30 is due to changing of carrier concentration brought by Y addition to Bi:2212 system. This leads to electronic or chemical inhomogeneity in the charge reservoir layer (BiO/SrO),


258

A. Sedky

and supplies the charge carriers to the CuO2 planes through which the actual supercurrent is believed to flow [54-56]. The reduction of these parameters at Y = 0.50 is due to enhancement of anisotropy along with the decrease of coupling between the CuO 2 planes. This is of course helps for producing weak links inhibiting the flow of supercurrent, and reduces the values of Bc and Jc. This is also consistent with the unusual values of OD for the Y = 0.50. Similar behavior is reported for the critical currents of Bi:2212 doped by Er, Fe and Ni at Cu site [57]. This is explained by increasing the pinning force density through the weak-link, and shifts the magnetic irreversibility line towards higher field values.

Figure 4(c). Coherence lengths versus Y content for Bi2Sr2Ca YxCu2Oy samples.

Figure 5(a). Critical magnetic fields versus Y content for Bi2Sr2Ca YxCu2Oy samples.


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 259

Figure 5(b). Critical current density versus Y content for Bi2Sr2Ca YxCu2Oy samples.

However, the unusual behavior of the OPD by Y additions in Bi: 2212 system implies that Y occupies the crystal structure of Bi: 2212 and decreased the c-parameter, and also leads to an increase in the oxygen content [35]. This oxygen could be incorporated into the Bi-O layers, and consequently an increase in the covalence of BiO bonds may be obtained [58]. Furthermore, the reduction of c-parameter should be decreases anisotropy and improves the coupling between the CuO 2 planes, and consequently the crossover behavior is affected as compared to pure sample. The second reason can be understood when Ca is replaced by Y, and the precipitated Y and Ca ions react with Bi and Cu ions to form a secondary phase rich in Cu. This causes either changes in the carrier concentration or can lead to both electronic and chemical in homogeneity of charge reservoir layers adjacent to the CuO 2 planes through which the actual super-current is believed to flow. However, it is approved that these secondary phases leads to the formation of different kinds of structural point like defects in the order of nanometers, acts as vortex pinning centers and improved the critical current density Jc [59].

CONCLUSION The Fluctuation induced excess conductivity study based on Bi2Sr2CaYxCu2Oy superconducting samples is reported. The OPD is shifted from (3D) to (2D) for pure sample, but it is shifted from (1D/qusi-1D/2D) to (qusi-1D/qusi-2D) for the Y doped samples. Furthermore, the c-axis coherence length at 0 K, critical fields and critical current density at 0K are improved by Y addition up to 0.30 followed by a decrease with further increase of Y up to 0.50. It is believed that some of secondary phases rich by Cu


260

A. Sedky

may be formed and strengthens by Y addition, and leads to point defects in the order of nanometers. These defects act as a vortex pinning centers and improved the critical values of the field and current as compared to pure sample.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8]

[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

H. Maeda, Y. Tanaka, M. Fukutomi, T. Asano, Jpn. J. Appl. Phys.27, 2, L209 (1988). C. A. M. dos Santos, S. Moehlecke, Y. Kopelevich and A. J. S. Machado, Physica C 390, 21 (2003). X. G. Lu, X. Zhao, X. J. Fan, X. F. Sun, W. B. Wu and H. Zhang, Appl. Phys. Lett. 76, 3088 (2000). A. Y. Ilyushchkin, T. Yamashita, L. Boskovic and I. D. R. Mackinnon, Supercond. Sci. Technol. 17, 1201 (2004). Sedky, Physica C 468, 1041 (2008). A. Sedky, J. of Physics and Chem. of Solids 70, 483 (2009). Sedky, J. Alloys and Compounds 499, 238 (2010). D. M. dos Santos, G. S. Pinto, B. Ferreira and A. J. S. Machado, Physica C 354, 388 (2001); A. D. M. dos Santos, S. Moehlecke, Y. Kopelevich and A. J. S. Machado, Physica C 390, 21 (2003). R. P. Aloysius, P. Guruswamy and U. Syamaprasad, Physica C 426-431, 556 (2005). K. Yamanaka, A. Suzuki, M. Suzuki and X. G. Zheng, Physica C 357-360, 237 (2001). M. Yilmazlar, H. A. Cetinkara, M. Nursoy, O. Ozturk and C. Terzioglu, Physica C 442, 101(2006). S. M. Khalil, J. Phys. Chem. Solids 64, 855 (2003). R. Singh and D. R. Sita, Physica C 312, 289(1999). R. Singh and D. R. Sita, Physica C 296, 21 (1998). Xuefeng Sun, Xia Zhao, Wenbin Wu, Xiaojuan Fan, Xiao- Guang Li and H. C. Ku, Physica C 307, 67(1998). R. P. Aloysius, P. Guruswamy, and U. Syamaprasad, Supercond. Sci. Technol. 18, L1 (2005). V. G. Prabitha, A. Biju, R. G. Abhilask Kumar, P. M. Sarun, R. P. Aloysius and U. Syamaprasad, Physica C 433, 28 (2005). Biju, R. P. Aloysius and U. Syamaprasad, Physica C 440, 52 (2006). Biju, P. M. Sarun, R. P. Aloysius and U. Syamaprasad, Materials Res. Bulletin 42, 2057 (2007).


Fluctuation Induced Excess Conductivity of Bi2Sr2CaYxCu2Oy Superconductors 261 [20] Biju, P. M. Sarun, R. P. Aloysius and U. Syamaprasad, J. Alloys Compound 431, 49 (2007). [21] A. K. Ghosh, S. K. Bandyopadhyay, P. Barat, Pintu Sen and A. N. Basu, Physica C 255, 319 (1995). [22] A. K. Ghosh, and A. N. Basu, Supercond. Sci. Technol. 13, 343 (2000). [23] Nawazish Ali Khan, Najmul Hassan, Sana Nawaz, Babar Shabbir, Sajid Khan and Azar A. Rizvi, J. Applied Physics 107, 083910 (2010). [24] A. K. Ghosh, S. K. Bandyopadhyay, and A. N. Basu, Mod. Phys. Lett. B 11, 1013 (1997). [25] D. S. Fisher, M. P. A. Fisher and D. A. Huse, Phys. Rev. B 43, 130 (1991). [26] F. Vidal, J. A. Veira, J. Maja, J. J. Ponte, F. G. Alvarado, E. Mordan, J. Amador and C. Cascales, Physica C 156, 807 (1988). [27] M. Mumtaz, S. M. Hasnian, A. A. Khurram and Nawazish A. Khan, J. Applied Physics 109, 023906 (2011). [28] Esmaeili, H. Sedghi, M. Amniat-Talab and M. Talebian, Eur. Phys. J. B 79, 443 (2011). [29] K. Semba, A .Matsuta and T. Ishii, Phys. Rev. B49, 10043 (1994). [30] P. Mandal, A. Poddar, A. N. Das, B. Ghosh and P. Choudhary, Physica C 169,43 (1990). [31] J. J. Wnuk, L. W. M. Schreurs, P. J. T. Eggenkamp and Van Der Linden Physica B 165-166, 1317 (1990). [32] M. O. Mun, S .I. Lee, S. H. S. Salk, H. J. Shin and M. K. Joo, Phys. Rev. B 48,6703 (1993). [33] Poddar, P. Mandal, A. N. Das, B. Ghosh and P. Choudhary, Physica C 161,567 (1989). [34] Natsuki Mori, J. A. Wilson and H. Ozaki, Phys. Rev. 45, 10633 (1992). [35] Sedky, J Supercond Nov Magn 29,1475 (2016). [36] Das et al. J. Physique 15, 623 (1995). [37] W. Anderson and Z. Zou, Phys. Rev. Lett. 60,132 (1988). [38] L. G. Aslamazov, A. I. Larkin, Phys. Lett. A 26, 238(1968); Sov. Phys. Solid State 10, 875 (1968). [39] W. E. Lawrence and S. Doniach, S. Proc. 12th Int. Conf. Low Temp. Phys. Kyoto, 1970 ( Edited by E. Kanada), P. 361, Keigaku, Tokyo ( 1971). [40] K. Gosh, S. K. Bandyopadhyay and A. N. Basu, J. Appl. Phys. 86, 3247(1999). [41] A. K. Ghosh, S. K. Bandyopadhyay, P. Barat, Pintu Sen and A. N. Basu, Physica C 264, 255 (1996). [42] A. Sedky, J. Low Temp. Phys. 148, 53 (2007). [43] M. V. Ramallo, C. Torron and F. Vidal, Physica C 230.97 (1994). [44] Baraduc and A. Bazdin, Phys. Lett. A, 171, 408 (1992). [45] S. Ravi and V. Seshu Bai, Solid State Commun 83,117 (1992).


262

A. Sedky

[46] B. D. Weaver, E. M. Jackson, G. P. Summers and E. A. Jackson, Phys. Rev. B 46, 1134 (1992). [47] J. A. Veira, J. Maza, F., J. Vida, Phys. Lett. A 131,310 (1988). [48] P. Mandal, A. Poddar, A. N. Das, J. Phys: Condensed Matter 6, 5689 (1994). [49] S. R. Ghorbani and M. Homaei, Modern Physics Letters B 25, 23, 1915 (2011). [50] Y. Petrovie, R. Fasano, M. Lortz, M. Dcrous, M. Potel and R. Cheriel, Physica C 460-462, 702 (2007). [51] Sedky, Journal of Magnetism and Magnetic Materials 277, 293 (2004). [52] P. Mandal, A. Poddar, B. Ghosh and P. Choudhary, Phys. Rev. B, 43,16, 13102(1991). [53] Matsuda, K. Kinoshita, T. Ishii, H. Shibata, T. Watanabe and T. Yamada, Phys. Rev. B 38, 2910 (1988). [54] R. P. Aloysius, P. Guruswamy, and U. Syamaprasad, Supercond. Sci. Technol. 18, L1 (2005). [55] Biju, P. M. Sarun, R. P. Aloysius and U. Syamaprasad, Materials Res. Bulletin 42, 2057 (2007). [56] Biju, P. M. Sarun, R. P. Aloysius and U. Syamaprasad, J. Alloys Compound 431, 49 (2007). [57] G. IIonca, T. R. Yang, A. V. Pop, G. Stiufiuc, R. Stiufiuc and C. Lung, Physica C 388-389, 425 (2003). [58] V. P. S. Awana, L. Menon, S. K. Mailk, Phys. Rev. B 51, 9379 (1995). [59] M. R. Koblischka, A. J. J. van Dalen, T. Higuchi, S. I. Yoo and M. Murakami, Phys. Rev. B 58, 2863 (1998).




Chapter 12

SUPERCONDUCTING MOTORS AND GENERATORS Jean Lévêque*, Kévin Berger and Bruno Douine Group of Research in Electrical Engineering of Nancy (GREEN), University of Lorraine, Vandœuvre-lès-Nancy, France

ABSTRACT The story of superconducting machines began in the 1960s. The first machine made in 1967 was a fully superconducting alternator using NbTi wires. Then, High Temperature Superconductors (HTS) were discovered and this was the starting point of many new studies on superconducting machines. This chapter reviews various topologies of superconducting motors and generators using HTS published in the literature. It begins with a brief presentation of the operating principle of electric motors. Then follows a striking description of the various realizations listed by machine typology. Some of these machines are totally innovative compared to conventional ones and their operating principle is strictly only related to the presence of superconducting materials.

Keywords: bulk superconducting materials, superconducting motors, superconducting wires and coils

1. INTRODUCTION Superconducting materials used in large-scale applications have the potential to reduce the overall size of the devices. In particular, the use of High Temperature Superconducting (HTS) materials motor applications of large scale may lead to huge *

Corresponding Author Email: [email protected].


264

Jean Lévêque, Kévin Berger and Bruno Douine

savings in size and losses compared to classical machines: increase of efficiency, specific power and torque. The considered powers are typically around some megawatts for applications like marine propulsion, embedded systems, windmill (Abrahamsen et al. 2012; Thongam et al. 2013; Karmaker et al. 2015; Kirtley, Banerjee, and Englebretson 2015; Yanamoto et al. 2015). This chapter presents recent developments in superconducting motors/generators worldwide and gives some data concerning constructed and tested prototypes. Many structures of superconducting motors have been studied. It would be time consuming and tedious to compile an exhaustive review. So, we have chosen to present some selected structures while skipping other ones. We will therefore discuss the synchronous, homopolar, claw pole, trapped flux and flux barriers machines. This chapter begins by a very short presentation of the structure of an electric engine. Electric motors are very old devices dating back to the 19 th century. The operating principle has remained unchanged since their discovery. Motors and generators are always composed by a fixed part and a rotating part. In a motor, two sets of coils called inductors and armature create two rotating fields at the same speed. The combination of these rotating fields creates a torque. This torque is proportional to the product of each of the rotating fields and the volume of the machine. In a generator, the inductor creates a rotating field. The armature is then submitted to induced currents and voltages at the same frequency of the rotating field. It is usual to introduce iron cores in the coils to increase the magnetic flux density produced and therefore the torque. The rotating fields can be created using two possible ways: either an electromagnet or a permanent magnet that rotates about its axis, or a set of coils feds by phase-shifted currents. This is depicted in Figure 1. The inductor and/or armature may be superconducting. Generally, the inductor rotates and the coils of the armature are fixed, but the system is fully reversible.

Figure 1. Illustration of magnetic rotating fields in an electrical motor.


Superconducting Motors and Generators

265

2. SYNCHRONOUS MOTOR 2.1. Introduction The majority of industrial designs are synchronous motors with radial flux. Machines of this type represent the machines for which we have a lot of data allowing an analysis based on the viewpoint of a superconducting machine. These superconducting machines are based on a design similar to conventional synchronous machines with salient poles and radial flux. The rotor consists of a superconducting excitation winding, while the armature of the stator is formed of a conventional copper armature. It is also possible to envisage a superconducting armature. On the other hand, all superconducting machines have been designed for “marine” or “windmill” applications and are, therefore, slow. So all our references will be related to this type of application. In a recent paper (Karmaker et al. 2015), a comparison is made between induction, permanent magnet and superconducting motors. The selected motor for comparison is an induction motor for an US warship. The respective data are summarized in Table 1. The comparison of three kinds of motors for ship propulsion is given in Figure 2. The authors conclude on the interest of each technology. The efficiency is calculated for a full load and are closed for all the system. The power density of a superconducting motor is more than two times better than an induction motor and a quarter better than a conventional permanent magnet motor. The losses in a conventional (1FJ4 801) and a superconducting motor (HTS II) produced by Siemens are presented in Figure 3. We can see that the losses are almost three times lower in the superconducting motor for the same output power. For these reasons, a general trend is about the studies of synchronous superconducting motor. Table 1. Specification of the conventional induction motor for comparison (Karmaker et al. 2015) Parameters

Values

Power

5000 HP

Voltage

590 V

Frequency

60 Hz

Turn

1791 rpm



266

Figure 2. Comparison of three kinds of motors for ship propulsion (Karmaker et al. 2015).

Figure 3. Comparison of losses of a conventional (1FJ4 801) and a superconducting motor (HTS II), obtained from Siemens experiments. (Nick, Grundmann, and Frauenhofer 2012).

2.2. Armature Two different designs are possible for the armature, either with a ferromagnetic yoke or without ferromagnetic yoke. In the case of conventional copper windings with iron teeth, the induction in the air gap is of the order of 1.2 Tesla. Another type of winding called “winding in the air gap” can be used to produce superconducting motors of a few megawatts. The aim is to obtain high inductions in the



267

gap, of the order of 2 T. This armature is composed of non-magnetic teeth made of composite material which makes it possible to keep the conductors made of copper and a magnetic crown for closing the external flux. Some photographs of the armature of the 400 kW motor by Siemens are shown in Figure 4.

Figure 4. Armature of the 400 kW-1500 rpm motor from Siemens.

These armature coils are subjected to alternating magnetic fields which generate additional losses: losses due to eddy currents and losses by circulation currents. These losses can be reduced by an appropriate choice of the size of the conductors (Litz wire) and twisting. These losses are extremely difficult to calculate and to predict, but it is possible to give orders of magnitude (Schuisky 1960; Lipo 2013): • •

Synchronous machines with salient poles: 0.1-0.2% Machines with smooth poles: 0.05-0.15%

2.2. Inductor The different designs of the inductor can be with or without iron to create the magnetic flux. Inductors without a ferromagnetic core are called “ironless rotor”. In this case, a non-magnetic structuring material is used instead of iron, generally, this material is based on a fiberglass composite, and it is used as a support for the superconducting coils. The use of iron in the inductor, which serves as a magnetic circuit to channel the magnetic flux, enables the amount of superconducting material required to be reduced.


268


Figure 5. Superconducting coils of the inductor of the 400 kW-1500 rpm motor from Siemens.

Two ways are possible for rotors with a ferromagnetic core, either this core is at cryogenic temperature (“cold-cored rotor”) or placed at ambient temperature (“warm iron-cored rotor”). For heavy-duty machines larger than 10 MW, which have a large volume and a large rotor mass, the use of iron leads to some problems concerning the cooling time. The cooling down of the system or its warmup to ambient temperature can take up to several months. In Figure 5, a photograph of the superconducting coils of the inductor of the Siemens 400 kW motor is shown.

3. SOME REALIZATIONS OF SYNCHRONOUS MOTORS In this part, some important projects, from Japan, Korea, United States and Germany are reported.

3.1. Kawasaki Heavy Industries (KHI) A group composed of Kawasaki Heavy Industries, Sumitomo, Electric Industries, TUMSAT, Yokohama, Sophia and Niigata University and National Maritime Research Institute designed and build a synchronous motor. The coils, made in DI-BSCCO are cooled down to 30 K by conduction cooling using helium gas. These coils generate a field up to 5 T, therefore, the inductor is completely without iron parts. The armature is made with copper wire and is also ironless to avoid too many losses in the iron teeth. In Figure 6, a photograph of a 3 MW motor from this group is presented. The results of the load tests of this 3 MW motor are given in Table 2.



269

Figure 6. 3 MW motor from KHI group (Yanamoto et al. 2015).

Table 2. Results of the load tests of the 3 MW KHI motor (Yanamoto et al. 2015) Speed 160.3 rpm 160.3 rpm 160.4 rpm

Torque 120.6 kN/m 150.6 kN/m 179.9 kN/m

Shaft output power 2.024 MW 2.527 MW 3.020 MW

This group, which has developed motors from 1 MW to 3 MW, is actually working on a 20 MW project. The main data of this recent project are summarized in Table 3. For this 20 MW motor, a thermosiphon system is used to cool down the coils using a mixture of Helium and Neon with a refrigerator that is completely integrated into the rotor. Table 3. Specifications of the 20 MW motor from KHI (Yanamoto et al. 2015) Parameters Rated power Rated speed Number of poles Number of slots Rated voltage of armature coil Rated current of field coil Field coil cooling Field coil operating temperature Outer diameter of back yoke Bearing distance Volumetric torque density

Values 20 MW 90 rpm 24 216 5200 V 200 A He & Ne gas 37 K 4m 3m 56 kN/m3


270


3.2. Doosan Heavy Industries In South Korea, Doosan Heavy Industries in collaboration with KERI (Korea Electrotechnology Research Institute) has developed a 2-pole, 1 MW generator rotating at a speed of 3600 rpm. The data of this generator are given in Table 4. The winding of the inductor is made with a BSCCO tape, cooled with liquid neon at 30 K. However, in order to simplify manufacturing, the armature is made with directly cooled hollow conductors, instead of using the Litz wire to reduce eddy current losses. This generator is mainly used for industrial applications: pump drives and power generation. Table 4. Specification of the 1 MW motor from Doosan (Baik and Park 2016) Parameters Rated power Rated speed Number of pole Number of slot Rated voltage of armature coil Rated current of field coil Field coil cooling Field coil operating temperature Size

Values 1 MW 3600 rpm 2 36 3300 V 150 A He & Ne gas 30-35 K 1.2 m x 2.4 m (diameter x length)

Figure 7. Photograph of the 1 MW motor from Doosan (Baik and Park 2016).



271

This motor is presented in Figure 7. There is no iron pole in the rotor, just a back iron for the armature, and no iron teeth. Therefore, the synchronous reactance is very low, several times smaller than for a classical motor. This leads to a smaller load angle and consequently a higher pull-out torque. Therefore, this machine is more stable than a conventional one. Through experimental values recorded, the authors have theoretically studied the power factor, the load angle and the efficiency of this motor. Efficiency versus output power for different terminal voltage is presented in Figure 8. Only copper losses are taken into account. The higher value of the efficiency at the lower voltage of 2900 V is due to a load factor near unity.

Figure 8. Efficiency versus output power for the 1 MW motor from Doosan (Baik and Park 2016).

3.3. AMerican Super Conductors (AMSC) Several research programs have been devoted to this machine topology in the United States. The most powerful engine was built by AMerican Super Conductors (AMSC) and Northrop Grumman in partnership with the U.S. Navy’s Office of Naval Research (ONR) program (Gamble, Snitchler, and MacDonald 2011). This ship propulsion engine has a power of 36.5 MW at 120 rpm, and it has been tested to full load successfully, as shown in Figure 9. This engine worked well, and compared to conventional engines, a gain of compactness (volume) more than 50% is obtained. The main characteristics of the engine are given in Table 5. The inductor of this motor is made using BSCCO wire. Finally, this motor is three time lighter than a conventional one.



272

Table 5. Specification of the 36.5 MW superconducting motor from AMSC Parameters Rated power Rated speed Number of poles Number of phases Rated voltage of armature coil Field coil cooling Field coil operating temperature Size Weight

Values 36.5 MW 120 rpm 16 9 6000 V He gas 30 K 3.4 m x 4.6 m x 4.1 m 75 Tons

Figure 9. Photograph of the 36.5 MW superconducting motor from AMSC (Gamble, Snitchler, and MacDonald 2011).

3.4. Siemens In Europe, the first High Temperature Superconducting machine was built by Siemens AG in Germany (Nick et al. 2002). This machine of 380 kW was made in a volume equivalent to that of a conventional machine of 260 kW, the inductor winding was designed with superconducting BSCCO tapes cooled down to 30 K using liquid Neon. The machine has been tested successfully in different operating modes: motor



273

powered via an inverter, and a generator synchronized to the network for half a year continuously. Encouraging results have been obtained, which led Siemens to design a generator ten times more powerful (Nick et al. 2007). This machine was completed in 2005, it is intended primarily for power generation applications within ships. Another project started in 2008 at Siemens (Nick et al. 2010) for the construction of a slow speed motor for marine propulsion, powered by a variable frequency inverter. This project is currently under test. The setup required a significant amount of superconducting BSCCO tape, estimated to be about 45 km length. The main specifications are summarized in Table 6 and a photograph of the motor is given in Figure 10. Table 6. Specification of the slow speed 4 MW motor from Siemens (Nick, Grundmann, and Frauenhofer 2012) Parameters Rated power Rated speed Rated voltage of armature coil Rated current Power factor Efficiency Weight

Values 4 MW 120 rpm 3100 V 775 A 1.0 96.2 % 37 tons

Figure 10. Photograph of the superconducting 4 MW motor from Siemens (Nick, Grundmann, and Frauenhofer 2012).



274

As for the previous generator and motor designed and build by Siemens, the cooling system is made using a thermosiphon with liquid Neon. This system uses four generators of the type Gifford McMahon AL 325. Each of them could remove 120 W at the operating temperature. The weight of this machine is lower than a conventional one. 25% is gained on the weight and 1.5% concerning the efficiency.

3.5. IHI Corporation In general, this axial magnetic flux machine topology is considered to be more compact than radial magnetic flux machines, even for conventional copper machines. One of the specific problems posed by this type of machine is the axial forces that must be maintained. For these reasons, the Japanese company IHI devotes an important effort to the development of this structure. In this case, it is a conventional permanent magnet motor with a superconducting armature. The Japanese Frontier Research Group, in close collaboration with IHI, has developed the most powerful engine cooled to 77 K: 400 kW-250 rpm, with DI-BSCCO ribbons cooled by liquid nitrogen (Okazaki, Sugimoto, and Takeda 2006). The group is made up of the following industrial partners and universities: Fuji Electric Systems Co., Ltd./Hitachi, Ltd./IHI Corporation/Nakashima Propeller Co., Ltd./Niigata Power Systems Co., Ltd./Sumitomo Electric Industries, Ltd./Taiyo Nippon Sanso Corporation/University of Fukui (Prof. Sugimoto). The main specifications of the motor are given in Table 7. A photograph and a sketch of the motor are shown in Figure 11, and the test bench is presented in Figure 12. The AC losses are principally located in the cryogenic parts, and not only in the HTS coils. Another important part of losses is located in the magnetic parts of this motor. Table 7. Specifications of the 400 kW axial flux motor from IHI Group (Oota and Fukaya 2016) Parameters Power Current supply Speed Size Weight

Values 400 kW 940 A 250 rpm 1.2 m x 0.8 m (diameter x length) 4.4 Tons



275

Figure 11. Photograph (a) and sketch (b) of the 400 kW axial magnetic flux motor from IHI Group (Oota and Fukaya 2016).

Figure 12. View of test bench of the 400 kW axial magnetic flux motor from IHI Group (Oota and Fukaya 2016).

3.6. ULCOMAP European Project In France, the first HTS machine was manufactured by Converteam Nancy in 2008 (Rezzoug, Lévêque, and Douine 2012), depicted in Figure 13. This 250 kW-1500 rpm demonstrator was realized within the framework of the ULCOMAP European project (ULtra-COmpact MArine Propulsion), which brought together several university and industrial partners (Enel, Futura composites, Zenergy power, Werkstoffzentrum, Converteam motors Nancy, GREEN Laboratory, Silesian University of Technology). This project aims to demonstrate the compactness gain of HTS motors compared to



276

conventional machines, for use in marine applications. This machine was made and tested at full load (Müller 2007). The main specifications of this motor are given in Table 8. Table 8. Specification of the 250 kW HTS motor realized by Converteam Nancy (Müller 2007) Parameters Power Voltage Supply Current supply Field coils current Speed Cooling system HTS wire Reactance d-axis Reactance q-axis

Values 250 kW 380 V 360 A 30 A 1500 rpm 30 K, Ne BSCCO 2223 0.22 pu 0.10 pu

Figure 13. Photograph of the 250 kW HTS motor realized by Converteam Nancy.

4. CLAW POLE MOTOR This part begins with a general presentation of the claw motors, and then we will have a closer look to applications of these superconducting machines.



277

4.1. Basic Principle and Description The rotating inductor has an excitation winding, which consists of a single coil, a single solenoid, fed to a brush-ring system. This coil is placed between discs with claws, allowing to create the poles. These discs with claws are made of a solid ferromagnetic material. Generally, the number of poles is 12, or even 16 for some models of significant powers. The stator is of classic design. This type of inductor is represented in Figure 14. In Figure 14, the central solenoid and the discs with the claws are shown. The disk behind the solenoid is defined arbitrarily to be the north pole of the coil, so that the disk below corresponds to the South Pole. A series of claws of different polarities are alternately attached to the north and the south poles. A spatial variation of the magnetic flux density is then performed with an inductor, which is easy to be manufactured. One of the advantages of this type of inductor is that in this way it is easy to create a large number of poles. Among the disadvantages are the claws, which are made of solid material and, therefore, the place of eddy currents losses. Due to its salience, this topology of motor is not convenient at too high speeds, and leads to huge aerolic losses. This type of motor is particularly used in the automotive industry for its simplicity of manufacture and robustness. It is also commonly used in low power generating sets.

Figure 14. Illustration of a claw pole motor.

4.2. Superconducting Claw Pole Motors Superconducting claw motors have been studied more specifically in Japan for automotive applications. This motor structure is well suited of superconductors because its inductor is a simple solenoid. Moreover, in the present version, the solenoid is fixed,



278

and only the claws rotate, which enables to simplify the required cryogenic parts. This system is presented in Figure 15. The most successful realization with this machine is certainly the one found in the framework of the development of an electric vehicle. A team of Sumitomo Electric motorized a Toyota car. Some characteristics of this engine fitted to a vehicle are given in Table 9 (Oyama et al. 2008).

Figure 15. Illustration of the components of a superconducting claw pole motor.

Table 9. Information concerning the superconducting claw pole motor from Sumitomo Electric (Oyama et al. 2008) Parameters Superconducting wire Coils size

Cooling Voltage supply Current supply Size Weight Power

Values BSCCO Sumitomo Internal radius: 186 mm External radius: 210 mm Length: 40 mm LN2 144 V 500 A 266 mm x 370mm (diameter x length) 110 kg 30 kW @ 2200 rpm

The voltage and power of this machine are closed to the one used in electrical cars. The cooling system is linked directly to the motors as shown in Figure 16. The superconducting coils are located in a bath of liquid nitrogen. On the top of the motor, a tank of liquid nitrogen is placed with the refrigerator in a closed loop. Therefore,



279

no refilling of liquid nitrogen is required. The distribution of the temperature in the motor is given in Figure 17.

Figure 16. Photograph of the superconducting claw pole motor including its cooling system (Oyama and Shinzato 2016).

Figure 17. Distribution of the temperatures in the superconducting claw pole motor from Sumitomo Electric (Oyama and Shinzato 2016).

The integration of this superconducting claw pole motor is shown in Figure 18. The power source, lead acid batteries, are located in the trunk of the car, see Figure 19.



280

Figure 18. Photograph of the integration of the superconducting claw pole motor (Oyama et al. 2008).

Figure 19. Photograph of the car where the superconducting claw pole motor has been integreted (Oyama et al. 2008).

Table 10. Vehicle performances (Oyama and Shinzato 2016) Parameters Torque Power Speed Operating range

Values 136 N.m @ 1540 rpm 30 kW @ 2200 rpm 80 km/h 36 km

The vehicle performances are given in Table 10. It is also important to notice that this was the first and unique implementation of a HTS motor in an application with real-size tests.

5. HOMOPOLAR MOTOR This concept of machine was discovered in 1831 by the British scientist Michael Faraday. The principle of operation is simple, see Figure 20. A conducting disk starts to



281

rotate when submitted to a constant magnetic field B, oriented axially, and simultaneously traversed by a continuous electric current I oriented radially. The resulting torque is proportional to the product I  B. For a more recent machine topology, the use of superconducting windings for the solenoids makes it possible to produce magnetic fields of several Tesla, which makes it possible to realize compact machines. It gives to this structure a renewed of interest, especially for use in marine propulsion applications (Gubser 2003). This topology is easy to implement, the rotor can be solid and the cold part fixed. The main problem of this structure is the feeding by a rings-brush system to transfer currents of high intensity. A Low Temperature Superconducting (LTS) homopolar generator was the first superconducting machine tested in situ a US Navy Jupiter II class ship (Stevens and Cannell 1981). It was a generator linked to a gas turbine and rotating at 19500 rpm. This represents, therefore, the first realistic application of LTS in electrical engineering.

Figure 20. Principle of a homopolar motor.

Following this first tests, a realization of this type of electrical machine using HTS materials (BSCCO 2223) dates back to 1996. The Naval Surface Warfare Center (Annapolis-USA), together with the Naval Research Center, have developed a 143 kW11 700 rpm motor cooled by liquid helium (Superczynski and Waltman 1997). General Atomics has a research program that aims to promote this structure for marine propulsion. After studying some low-power prototypes (Figure 21), this company built the largest HTS homopolar demonstrator currently known (Thome, Bowles, and Reed 2006). This machine was designed for a power of 3.7 MW and is shown in Figure 22. One of the most critical problems of homopolar motors is the slip ring and brush system, which has to transmit very high currents at a high speed. The Guinea research & development Company located in Australia develops liquid-metal ring systems and homopolar motors as shown in Figure 23 (Fuger et al. 2015). The main characteristics of this motor are summarized in Table 11.


282


Figure 21. Illustration of a homopolar motor concept from General Atomics (Thome et al. 2002).

Figure 22. Photograph of the 3.7 MW HTS homopolar motor from General Atomics (Thome et al. 2002).

Figure 23. Photo of the 200 kW HTS homopolar motor from Guinea (Fuger et al. 2015).



283

Table 11. Specifications of the 250 kW HTS homopolar motor from Guinea (Fuger et al. 2015) Parameters Power Speed Current Voltage

Values 200 kW 3 600 rpm 20 000 A 10 V

It should be emphasized that the development of these engines is still far from completion.

6. MOTORS WITH HTS BULKS AS MAGNETIC SCREENS Like variable reluctance machines, these structures utilize the shielding properties of superconducting materials to modulate and concentrate a magnetic field produced, for example, by solenoids. These machines have atypical rotor topologies where the main technological limitation to their realization remains the design of HTS pellets of large sizes with high capacities of shielding.

6.1. Magnetic Flux Modulation Motor This first atypical structure, proposed by the GREEN Laboratory, is composed of two solenoids of superconducting wires fed by inverse currents; the resulting magnetic fields will therefore be in opposition and lead to a total field with a strong radial component. Between these two windings are inserted HTS pellets, YBCO type, which by their properties to expel the magnetic flux density (Lenz’s law), will modulate in space the magnetic flux density around the circumference. This variation on the contour of the inductor evolves between a minimum value close to zero behind the HTS bulks and a maximum value in between. The magnetic flux density created is then directly proportional to the currents in the solenoids. The two successive PhD thesis of P. Masson and E. H. Ailam, led by A. Rezzoug and co-supervised by J. Lévêque and D. Netter, allowed the design of a superconducting synchronous motor with four pairs of poles. A first realized prototype of inductor validated the principle of this structure (Masson et al. 2003). The design and tests of a motor based on this principle was realized at GREEN laboratory (Ailam et al. 2007). This topology allows to consider high magnetic flux density in the air gap and therefore to obtain high mass torque. On this principle, axial flux topologies could be envisaged (Masson, Tixador, and Luongo 2007).


284


Figure 24. Schematic design of the magnetic flux modulation superconducting as proposed originally by GREEN.

6.2. Magnetic Flux Barrier Motor This second structure is firstly proposed by the GREEN laboratory (Alhasan et al. 2015; Alhasan, Lubin, Adilov et al. 2016; Alhasan, Lubin, Douine et al. 2016). The geometry remains composed of two superconducting solenoids, but fed in the same direction. Like Helmholtz's coils, these two solenoids generate magnetic flux density in the same direction. The variation of the magnetic flux density on the contour of the inductor is obtained by the addition of an inclined HTS bulk, arranged between these two windings. This HTS bulk acts as a magnetic screen. An iron core can be added in between the HTS bulk and the superconducting coils in order to strengthen the magnetic flux density in the air gap of the machine. A machine structure with one pair of poles is then obtained, and the schematic design is given in Figure 25. The operating principle could be assimilated to synchronous claw pole machines. Iron core Armature Bulk HTS Magnetic field

Magnetic field

Current

Flux line Back iron

Figure 25. Schematic design of the flux barrier motor as proposed originally by GREEN.



285

7. TRAPPED FLUX MOTOR One of the keys to obtain superconducting motors with high specific power is to optimize the magnetic trapped flux in massive superconducting materials to reach magnetic flux density greater than 2 T. Like the permanent magnets, these HTS bulks are used in axial magnetic flux machines (Miki et al. 2006; Sugimoto et al. 2008) or radial (Xian et al. 2011). However, the main difficulty lies in the magnetization process of these pellets confined in a cryogenic environment. Indeed, the coil used for the magnetization can be integrated inside the cryostat or in the ambient environment. Presently, the preferred method, in terms of space requirement and ease of implementation, remains the so-called “Pulse Field Magnetization.” Another way is to use a stack of HTS tapes which acts as a magnet. This allows to take advantage of the excellent superconducting properties of YBCO coated conductors. Currently, the works carried out in the field of HTS bulk magnets make it possible to consider trapped flux density by PFM close to several Tesla (Fujishiro et al. 2006; Gony et al. 2015; Patel et al. 2015). An original prototype of an axial magnetic flux density synchronous motor, developed by a group of universities in Japan, was tested with liquid nitrogen (Matsuzaki et al. 2007). (RE)BCO pellets are initially magnetized by the armature. Superconducting permanent magnets are then obtained and act as an inductor. The stator copper windings can be used as an armature of the electrical machine. The structure, with 8 HTS bulks on the rotor part and a double armature, was successfully tested. Another study on axial magnetic flux motor, yielded a power of 16 kW, with a double rotor, is composed of 16 HTS bulks, and a triple armature (Matsuzaki et al. 2007), which is depicted in Figure 26.

Figure 26. Axial trapped magnetic flux motor, with double rotor (Matsuzaki et al. 2007)].


286


Figure 27. Superconducting trapped flux motor realized by Cambridge University.

A superconducting trapped flux motor build by the University of Cambridge (Xian et al. 2011), uses two coil windings to magnetize a rotor covered with in total 75 YBCO pellets as shown in Figure 27. The armature, composed of six BSCCO pancake coils enables the engine operation. The inductor is magnetized out of the armature and then moves down into the latter after magnetization. The set of two coils is used to magnetize one by one each of the poles of the rotor by rotating it by 90°. In Figure 27, one can see the inductor with the magnetizing coils on the top of the motor and below this part the armature. A PFM process is used to magnetize the HTS bulks placed on the inductor (Huang et al. 2016).

CONCLUSION It appears that up to now many superconducting motors have been produced. They are not only prototypes, but already also industrial demonstrators. Their efficiency and even their reliability have been proved. The present achievements are mainly engines and generators for the navy and marginally, for automobiles. The main obstacle to their commercialization is the prohibitive price which is directly linked to the cost of the superconducting wire, despite their undeniable qualities. Indeed, these engines are about four to five times more expensive than conventional permanent magnet motors at the present time. A sufficient market demand is needed to push this technology forward and to convince the end users to enter this technology. Another difficulty is the availability and the reliability of HTS wires. Standardized products as superconducting wires and coils are also required for the industrial



287

applications. For motor manufacturers, the challenge is to industrialize the process of fabrication of a superconducting motor.

REFERENCES Abrahamsen, A. B., N. Magnusson, B. B. Jensen, and M. Runde. 2012. ‘Large Superconducting Wind Turbine Generators.’ Energy Procedia, Selected papers from Deep Sea Offshore Wind R&D Conference, Trondheim, Norway, 19-20 January 2012, 24 (January): 60–67. doi:10.1016/j.egypro.2012.06.087. Ailam, E. H., D. Netter, J. Leveque, B. Douine, P. J. Masson, and A. Rezzoug. 2007. ‘Design and Testing of a Superconducting Rotating Machine’. IEEE Transactions on Applied Superconductivity 17 (1): 27–33. doi:10.1109/TASC.2006.887544. Alhasan, R., T. Lubin, Z. M. Adilov, and J. Lévêque. 2016. ‘A New Kind of Superconducting Machine’. IEEE Transactions on Applied Superconductivity 26 (3): 1–4. doi:10.1109/TASC.2016.2531003. Alhasan, R., T. Lubin, B. Douine, Z. M. Adilov, and J. Lévêque. 2016. ‘Test of an Original Superconducting Synchronous Machine Based on Magnetic Shielding’. IEEE Transactions on Applied Superconductivity 26 (4): 1–5. doi:10.1109/ TASC.2016.2536785. Alhasan, R., T. Lubin, J. Lévêque, K. Berger, B. Douine, A. Rezzoug, S. Mezani, G. Didier, M. Hinaje, and D. Netter. 2015. ‘Study of a Superconducting Motor with High Specific Torque.’ In MEA 2015 More Electric Aircraft, 91. Toulouse, France: 3AF, SEE. https://hal.archives-ouvertes.fr/hal-01113838. Baik, S. K., and G. S. Park. 2016. ‘Load Test Analysis of High-Temperature Superconducting Synchronous Motors’. IEEE Transactions on Applied Superconductivity 26 (4): 1–4. doi:10.1109/TASC.2016.2530662. Fuger, R., A. Guina, D. Sercombe, J. Kells, A. Matsekh, K. Labes, T. Lissington, C. Fabian, and G. Chu. 2015. ‘Superconducting Motor Developments at Guina Energy Technologies’. In 2015 IEEE International Conference on Applied Superconductivity and Electromagnetic Devices (ASEMD), 362–63. doi:10.1109/ ASEMD.2015.7453613. Fujishiro, H., T. Tateiwa, A. Fujiwara, T. Oka, and H. Hayashi. 2006. ‘Higher Trapped Field over 5 T on HTSC Bulk by Modified Pulse Field Magnetizing’. Physica C: Superconductivity and Its Applications, Proceedings of the 18th International Symposium on Superconductivity (ISS 2005), 445 (October): 334–38. doi:10.1016/ j.physc.2006.04.077. Gamble, B., G. Snitchler, and T. MacDonald. 2011. ‘Full Power Test of a 36.5 MW HTS Propulsion Motor’. IEEE Transactions on Applied Superconductivity 21 (3): 1083– 88. doi:10.1109/TASC.2010.2093854.


288


Gony, B., K. Berger, B. Douine, M. R. Koblischka, and J. Lévêque. 2015. ‘Improvement of the Magnetization of a Superconducting Bulk Using an Iron Core’. IEEE Transactions on Applied Superconductivity 25 (3): 1–4. doi:10.1109/ TASC.2014.2373494. Gubser, D. U. 2003. ‘Superconducting Motors and Generators for Naval Applications’. Physica C: Superconductivity, Proceedings of the 15th International Symposium on Superconductivity (ISS 2002): Advances in Superconductivity XV. Part II, 392 (October): 1192–95. doi:10.1016/S0921-4534(03)01124-9. Huang, Z., H. S. Ruiz, Y. Zhai, J. Geng, B. Shen, and T. A. Coombs. 2016. ‘Study of the Pulsed Field Magnetization Strategy for the Superconducting Rotor’. IEEE Transactions on Applied Superconductivity 26 (4): 1–5. doi:10.1109/ TASC.2016.2523059. Karmaker, H., D. Sarandria, M. T. Ho, J. Feng, D. Kulkarni, and G. Rupertus. 2015. ‘High-Power Dense Electric Propulsion Motor’. IEEE Transactions on Industry Applications 51 (2): 1341–47. doi:10.1109/TIA.2014.2352257. Kirtley, J. L., A. Banerjee, and S. Englebretson. 2015. ‘Motors for Ship Propulsion’. Proceedings of the IEEE 103 (12): 2320–32. doi:10.1109/JPROC.2015.2487044. Lipo, Thomas A. 2013. Introduction to AC Machine Design. Madison, WI: University of Wisconsin. Masson, P. J., P. Tixador, and C. A. Luongo. 2007. ‘Safety Torque Generation in HTS Propulsion Motor for General Aviation Aircraft’. IEEE Transactions on Applied Superconductivity 17 (2): 1619–22. doi:10.1109/TASC.2007.898114. Masson, P., J. Leveque, D. Netter, and A. Rezzoug. 2003. ‘Experimental Study of a New Kind of Superconducting Inductor’. IEEE Transactions on Applied Superconductivity 13 (2): 2239–42. doi:10.1109/TASC.2003.813055. Matsuzaki, H., Y. Kimura, E. Morita, H. Ogata, T. Ida, M. Izumi, H. Sugimoto, M. Miki, and M. Kitano. 2007. ‘HTS Bulk Pole-Field Magnets Motor with a Multiple Rotor Cooled by Liquid Nitrogen’. IEEE Transactions on Applied Superconductivity 17 (2): 1553–56. doi:10.1109/TASC.2007.898488. Miki, M., S. Tokura, H. Hayakawa, H. Inami, M. Kitano, H. Matsuzaki, Y. Kimura, et al. 2006. ‘Development of a Synchronous Motor with Gd–Ba–Cu–O Bulk Superconductors as Pole-Field Magnets for Propulsion System’. Superconductor Science and Technology 19 (7): S494. doi:10.1088/0953-2048/19/7/S14. Müller, Jens. 2007. ‘Final Report ULCOMAP, Project Title: ULCOMAP, Contract N° G3ST-CT-2002-50337 ULCOMAP’. Nick, W., M. Frank, G. Klaus, J. Frauenhofer, and H. W. Neumuller. 2007. ‘Operational Experience with the World’s First 3600 Rpm 4 MVA Generator at Siemens’. IEEE Transactions on Applied Superconductivity 17 (2): 2030–33. doi:10.1109/ TASC.2007.899996.



289

Nick, W., M. Frank, P. Kummeth, J. J. Rabbers, M. Wilke, and K. Schleicher. 2010. ‘Development and Construction of an HTS Rotor for Ship Propulsion Application’. Journal of Physics: Conference Series 234 (3): 032040. doi:10.1088/17426596/234/3/032040. Nick, W, G. Nerowski, H. W. Neumüller, M Frank, P van Hasselt, J Frauenhofer, and F Steinmeyer. 2002. ‘380 KW Synchronous Machine with HTS Rotor Windings–– development at Siemens and First Test Results’. Physica C: Superconductivity 372 (August): 1506–12. doi:10.1016/S0921-4534(02)01069-9. Nick, Wolfgang, Joern Grundmann, and Joachim Frauenhofer. 2012. ‘Test Results from Siemens Low-Speed, High-Torque HTS Machine and Description of Further Steps towards Commercialisation of HTS Machines’. Physica C: Superconductivity and Its Applications, 2011 Centennial superconductivity conference – EUCAS–ISEC– ICMC, 482 (November): 105–10. doi:10.1016/j.physc.2012.04.019. Okazaki, T., H. Sugimoto, and T. Takeda. 2006. ‘Liquid Nitrogen Cooled HTS Motor for Ship Propulsion’. In 2006 IEEE Power Engineering Society General Meeting, 6 pp.-. doi:10.1109/PES.2006.1709647. Oota, Tomoya, and Atsuko Fukaya. 2016. ‘Axial-Gap Superconducting Synchronous Motors Cooled by Liquid Nitrogen’. Research, Fabrication and Applications of Bi2223 HTS Wires 1: 451. Oyama, H., T. Shintato, K. Hayashi, K. Kitajima, A. Ariyoshi, and T. Sawai. 2008. ‘Application of Superconductors for Automobiles’. SEI Tech. Rev., no. 67: 22–26. Oyama, H., and T. Shinzato. 2016. ‘Development of High-Temperature Superconducting DC Motor for Automobiles’. Research, Fabrication and Applications of Bi-2223 HTS Wires 1: 485. Patel, A., A. Baskys, S. C. Hopkins, V. Kalitka, A. Molodyk, and B. A. Glowacki. 2015. ‘Pulsed-Field Magnetization of Superconducting Tape Stacks for Motor Applications’. IEEE Transactions on Applied Superconductivity 25 (3): 1–5. Rezzoug, A., J. Lévêque, and B. Douine. 2012. ‘Superconducting Machines’. In NonConventional Electrical Machines, eds. A. Rezzoug and M. El-Hadi Zaim, John Wiley & Sons Inc., Chap. 4, pp. 191–255. Schuisky, W. 1960. Berechnung elektrischer Maschinen. [Calculation of electrical machines.] Wien: Springer. Stevens, H. O., and M. J. Cannell. 1981. ‘Active Superconductive Generator Development 400-Horsepower Generator Design.’ Propulsion and Auxiliary Systems Dept. David W. Taylor Naval Ship R&D Center, Annapolis, Maryland 21402. http://www.dtic.mil/docs/citations/ADA107513. Sugimoto, H., T. Morishita, T. Tsuda, T. Takeda, H. Togawa, T. Oota, K. Ohmatsu, and S. Yoshida. 2008. ‘Development and Test of an Axial Flux Type PM Synchronous Motor with Liquid Nitrogen Cooled HTS Armature Windings’. In Journal of


290


Physics: Conference Series, 97:012203. IOP Publishing. http://iopscience.iop.org/ article/10.1088/1742-6596/97/1/012203/meta. Superczynski, M. J., and D. J. Waltman. 1997. ‘Homopolar Motor with High Temperature Superconductor Field Windings.’ IEEE Transactions on Applied Superconductivity 7 (2): 513–518. Thome, R. J., W. Creedon, M. Reed, E. Bowles, and K. Schaubel. 2002. ‘Homopolar Motor Technology Development’. In IEEE Power Engineering Society Summer Meeting, 1:260–64 vol.1. doi:10.1109/PESS.2002.1043229. Thome, R. J., E. Bowles, and M. Reed. 2006. ‘Integration of Electromagnetic Technologies into Shipboard Applications’. IEEE Transactions on Applied Superconductivity 16 (2): 1074–1079. Thongam, J. S., M. Tarbouchi, A. F. Okou, D. Bouchard, and R. Beguenane. 2013. ‘Trends in Naval Ship Propulsion Drive Motor Technology’. In Electrical Power & Energy Conference (EPEC), 2013 IEEE, 1–5. IEEE. http://ieeexplore.ieee.org/ abstract/document/6802942/. Xian, W., Y. Yan, Weijia Yuan, R. Pei, and T. A. Coombs. 2011. ‘Pulsed Field Magnetization of a High Temperature Superconducting Motor’. IEEE Transactions on Applied Superconductivity 21 (3): 1171–1174. Yanamoto, T., M. Izumi, M. Yokoyama, and K. Umemoto. 2015. ‘Electric Propulsion Motor Development for Commercial Ships in Japan’. Proceedings of the IEEE 103 (12): 2333–2343.




Chapter 13

A REVIEW ARTICLE: COMPACT MAGNETIC FIELD GENERATORS CONTAINING HTS BULK MAGNETS COOLED BY REFRIGERATORS AND THEIR FEASIBLE APPLICATIONS Tetsuo Oka1,*, Jun Ogawa1, Satoshi Fukui1, Takao Sato1, Tomohito Nakano1, Kazuya Yokoyama2, Takashi Nakamura3, Hiroyuki Fujishiro4 and Koshichi Noto4,* 1

2

Faculty of Engineering, Niigata University, Niigata, Japan Faculty of Engineering, Ashikaga Institute of Technology, Ashikaga, Japan 3 Center for Sustainable Resource Science, RIKEN, Wakoh, Japan 4 Faculty of Science and Engineering, Iwate University, Morioka, Japan

ABSTRACT Various types of strong magnetic field generators with use of HTS bulk magnets and the small-scale refrigerators have been actually proposed and experimentally estimated for the feasible applications. The most characteristic feature of bulk magnets is regarded as compact and strong field generators with respect to their magnetic field spaces or their machine structures of themselves. Among them, the magnetic flux density of the face-toface magnetic poles reached 3.2 T and 4.4 T when they were activated by the pulsed field magnetization and the field cooling method, respectively. Furthermore, a very small bulk *

Corresponding author: [email protected].


292

Tetsuo Oka, Jun Ogawa, Satoshi Fukui et al. magnet generating 2.8 T and the system featured by the wide magnetic poles containing seven magnets have been constructed with use of ST pulse tube coolers in order to expand the feasible application areas. On the other hand, although the bulk magnets are characterized by their steep field distributions, we attempted to gain the uniform magnetic field space, and successfully achieved the uniformity of 385 ppm in 4 mm range by combining the deformed magnetic field profiles which generated between the face-to-face magnetic poles. The performances of these unique systems have exceeded those of the conventional electromagnet or any permanent magnets by far, demonstrating the superiority to other strong magnetic field generators.

Keywords: high temperature superconductor, bulk magnet, refrigerator, pulsed field, magnetic field trapping

INTRODUCTION The melt-processed high temperature superconducting materials (hereafter abbreviated as HTS) basically composed of RE(Y, Sm, Gd, Dy, Eu)-Ba-Cu-O have often been called the trapped-field magnets or bulk magnets when they capture the magnetic fields applied from outside under their superconducting states. The materials are capable of generating intense magnetic fields more than several T by applying either static or pulsed magnetic fields [1, 2]. People have expected the HTS devices to be utilized at 77 K, because of their high Tc property. However, since the handling of cryogen requires special techniques and does not suit the widespread industries, the authors have emphasized the importance of employing the refrigerators instead of the cryogen such as liquid nitrogen in order to keep the superconductivity when we intend to apply these materials to the practical industries. Furthermore, we all know that it is much effective to use the HTS materials in the temperature range lower than 77 K, because the superconducting properties, e.g., critical current density Jc values, are substantially enhanced [3]. In 1999, Ikuta et al. [4, 5] reported that the trapped field arose up to 9 T at 25 K, which is more than 5 times stronger than that at 77 K. Krabbes et al. [6, 7] and Tomita et al. [8] have shown the excellent trapped fields of 16.0 T at 24 K and 17.24 T at 29 K, respectively. Recently, Durrell et al. reported the highest field-trapping of 17.6 T at 26 K [9]. These records were achieved by the so-called field cooling magnetization (FCM) method in the static magnetic field generated by a superconducting solenoid magnet [10], in which the bulk magnets have to withstand the intense mechanical stress caused by the pinning force [11-13]. On the other hand, the pulsed-field magnetization (PFM) method has been developed as a simpler and easier operation than FCM [14-16]. Since the pinning effect under 77 K is enhanced with lowering temperature, we are well aware of the inferiority in field-trapping performances obtained by PFM to FCM [17-19] because of the flux pinning [20, 21] and the heat generation caused by the magnetic flux motion


A Review Article

293

during the process [22-24], Fujishiro et al. reported the world’s highest performance of 5.2 T through their extensive studies during the activation processes [25, 26]. The positions of the bulk magnets among other magnetic field generators have been discussed from the industrial point of view. The bulk magnet systems combined with the small-scale refrigerators are characterized by their intense and compact magnetic fields [27]. Actually, the authors have constructed various types of the strong magnetic field generators with use of bulk magnets and the small-scale refrigerators [28]. In this chapter, we discuss these bulk magnet systems, evaluate their characteristics in order to realize the industrialization of intense and compact magnetic fields, and finally introduce a novel way to obtain the uniform magnetic field space which would be available as strong field generators for the feasible analytic devices which, a priori, require homogeneous and stable magnetic fields.

EXPERIMENTAL PROCEDURE HTS Bulk Magnets In this paper, we deal with the melt-processed REBa2Cu3Oy bulk magnets containing fine RE2BaCuO5 grains, which were manufactured by Nippon steel and Sumitomo Metal Co., and Dowa Mining Co. The dimensions are 60 or 45 mm in diameter and 15 mm in thickness. The trapped magnetic field distribution of each sample at 77 K exhibits a fine conical shape. The bulk magnets were reinforced by putting the stainless-steel rings on to withstand the fractures due to the magnetic force during the activation processes [29].

Cooling Systems The small-scale Gifford-McMahon (GM)-cycle refrigerators or ST pulse tube coolers (both of them were made by AISIN SEIKI CO.) were employed to compose the systems. The bulk magnets were settled in the vacuum chambers equipped by the turbo molecular pumps and the rotary pumps to keep the bulk magnet in the thermally-insulated condition, and then cooled to 31 K and 64 K by the former and latter refrigeration devices, respectively. The nominal output/input powers of GM-cycle cooler and ST pulse tube cooler was reported as 15/1,000 W and 8/300W at 77 K, respectively [30].


294

Tetsuo Oka, Jun Ogawa, Satoshi Fukui et al.

Magnetizing Procedures In the paper, PFM and FCM methods are employed to exhibit the activation methods for the bulk magnets. The PFM is, we believe, the easiest among all possible magnetizing techniques. However, the flux penetration into the bulk magnets is sluggish due to the strong pinning forces [15, 23]. In addition, the flux motion during the penetration process causes local heating in the sample, raises the temperature, lowers the local Jc, and subsequently degrades the field trapping ability. The IMRA method (Iteratively magnetizing pulsed-field operation with reducing amplitudes) effectively enhances the trapped field [16]. The change of the magnetic field in the regions where the magnetic fluxes have already penetrated is restrained as if they would have been repelled away [31, 32]. As a result, the heat generation is effectively suppressed when further successive magnetic pulse is applied. The FCM process was conducted in a room temperature bore of the superconducting solenoid magnet (Japan Superconductor Technology Co.). The bulk magnet is directly cooled in it by a GM cooler without using liquid helium. The magnetic pole before activation were inserted into the bore, and the static fields of 3 - 6 T were applied in the normal conducting temperature range upper than Tc, and then the sample was cooled to the lowest temperature of each refrigerator. After that, the magnetic field was removed to zero by slow descending rate to avoid the heat generation [10].

Estimation of Generating Magnetic Fields The trapped field distributions after the PFM and FCM processed were measured in two ways of scanning a Hall sensor (F. W. Bell, BHA921, BHT921) just above the vacuum chamber and attaching it on the bulk surface in the chamber in order to trace the evolutions of the magnetic flux density during the PFM processes. In order to draw the distribution map, we scanned the sensor with every 2-mm pitch and 0.5-s intervals. The direction of thus measured magnetic field was parallel to the pole axis Bz in the open gaps between the magnetic poles [15].

RESULTS AND DISCUSSION Bulk Magnets Systems Activated by Pulsed Field Magnetization Figure (1a) shows the whole bulk magnet system [27]. A bulk superconductor is settled on the cold stage cooled by GM cooler. We need to prepare a vacuum pump to


A Review Article

295

maintain thermal insulation from outside and magnetizing devices to activate them. The magnetizing coil and condenser bank are taken away after the PFM operation, and the compactly-composed single magnetic pole system, which is generating the magnetic flux over 2 T from its magnetic pole surface, needs only the continuing drive of refrigerators to keep the magnet at the temperature ranges under the Tc. As shown in Figure (1b), one can easily carry the magnetic pole in hand, since the weight is designed to be less than 10 kg. The bulk magnet system is characterized as a compact and strong magnetic field generator. The authors have constructed various bulk magnet systems, which were designed and employed to the feasible industrial applications, as shown in Figure (2) [27]. The all-inone type magnet was built at first to aim the most compact field generator, which was utilized as a demonstrator to show the feasible intense field. Then, the field generator was separated to the pole piece and the compressor to adopt them to the magnetron sputtering cathode [33-35] and the liquid stirrer. In order to enhance the magnetic field performances, the face-to-face magnetic poles were designed to obtain the doubled magnetic field strength. The magnet has been employed to study the magnetic separation, and so on [36]. Magnetic flux emission Magnetic pole surface

Magnetizing pulse coil

RE123 bulk Cold head

Condenser

Vacuum vessel

He compressor

HTS REBCO bulk magnet (in vacuum)

GM refrigerator in hand

Magnetic Flux Magnetic pole

Vacuum pump control Thermometer

Refrigerator Turbo molecular pump

(a)

Rotary pump

Vacuum pump (in the frame) Compressor

(b)

Gas-He tube

Figure 1. Elements of HTS bulk magnet system activated by PFM (a), and a view of compact magnetic field generator (b).

A novel face-to-face type high-field bulk magnet system has been constructed in order to achieve further strong magnetic field for feasible applications. Figure (1d) shows the face-to-face type bulk magnet system activated by FCM method operated at 6 T. The resultant magnetic field of 4.4 T was trapped in one side of the magnetic poles. On the other, when they were activated by 5-T FCM, they have formed the 3.98 T magnetic poles. The magnetic field measured at the center of the gap space reached 2.13 T between the magnetic poles. The intense magnetic field is used in the experiment such as magnetization of the permanent magnets in the dc brushless motor [37-39]. Furthermore,



296

the GM pulse tube refrigerator was employed to the R&D project for the NMR magnet in which the operation without any vibration was necessary [40]. GM single-pole (separated type)

GM single-pole

H

H

(a)

Magnetron sputtering Liquid stirrer

(b)

Face-to-Face type by PFM

Face-to-Face by Field Cooling GM refrigerator

REBCO bulk magnets

Vacuum valve

H Magnetic Flux

He compressor

(c)

Magnetic separation NMR magnet

(d)

Thermocontrol

He compressor

Magnetic separation Magnetizing device

Figure 2. Various magnetic field generators composed of GM coolers.

In order to enlarge the strong field space, a novel magnetic field generator with a pair of wide magnetic poles was constructed by using seven bulk magnets arranged in one plane of 150 mm in diameter, as is shown in Figure (3a). A pair of magnetic poles with a size of 230 mm in diameter was located face to face after each FCM process operated at 67 K with use of the ST pulse tube coolers [27, 28]. As shown in Figure (3b), the magnetic flux densities Bz of 0.89 T were obtained at the center of the gap distance between the magnetic poles. One sees seven distinctive peaks which reflected the positions of the bulk magnets embedded in each magnetic pole even when the gap of the magnetic poles was 34 mm. The maximum value of the total fluxes has reached 4.3 mWb, which corresponds to nearly four times as much as that obtained when the single bulk sample was magnetized in the temperature range obtained GM cooler.


A Review Article

297

Magnetic pole

Flux density Bz(mT)

700

ST Pulse tube cooler 7 bulk magnets

600 500 400 300 200 100 0 -100 0

100 50

100 Bulk magnet stage position (mm)

(a)

position (mm)

0 200

150

(b)

Figure 3. A photo of the multi-bulk arrangement with wide magnetic pole surfaces (a), and the magnetic field distribution Bz measured between the poles after FC magnetization (b). Bulk magnet (in the vessel)

Cold head ( in the vessel)

Magnetic pole

Vacuum pump

Magnetic field H

Sensors

Magnetic flux density (T)

3.0

Bz max = 2.78T 2.5 -3.0 2.0 -2.5 1.5 -2.0 1.0 -1.5 0.5 -1.0 0.0 -0.5 -0.5 -0.0

2.5 2.0 1.5 1.0

0.5 0.0 -0.5

Thermometer

ST pulse tube cooler

(a)

Voice coil-type motor

0 y-Position (mm)

0

40

(b)

70 x-Position (mm)

80

Figure 4. Small-scale HTS bulk magnet system composed of compact ST pulse tube cooler (a), and the performance Bz after activation by 5 T FC mode.

Figure (4) shows another small-scale bulk magnet with use of the ST pulse tube cooler. The magnetic pole contained a Gd123 bulk magnet with a size of 60 mm in diameter, which was directly attached to the bulk stage which was connected by a heat conduction rod made of copper to the cold head of the cooler. The whole system was capable of being placed on the ordinary office desks, since the weight was designed to be lighter than 20 kg. The temperatures of the bulk stage reached the lowest temperature of 59 K in 8 h. Since the ultimate temperature of the cooler was a bit higher than that of the


298


GM cooler, the FCM process was operated instead of PFM to avoid the heat generation in the bulk magnet. The pre-activated magnetic pole was inserted into the room temperature bore of the superconducting solenoid magnet. A static field of 5 T was applied in the higher temperature range than Tc, and then the bulk magnet was cooled to the lowest temperature which the refrigerator could reach. After the reduction of magnetic field to zero with a descending rate of 4.2 mT/s, the maximum trapped field was obtained as 2.78 T at the surface of the magnetic pole by scanning the Hall sensor [30, 41].

Magnetic Field Distribution of Face-to-Face Bulk Magnet Systems Figure (5a) shows the face-to-face magnet system containing a pair of Sm–Ba–Cu–O bulk magnets on the respective cold stages of the GM refrigerators [27, 28, 36]. One sees the pulse coil on the right-hand side on the equipment. After the PFM activation, the trapped field distribution Bz was measured by scanning a Hall sensor in the open 2 mm gap between the magnetic poles. After the activation process so-called IMRA method [16], the field trapping property is generally affected by the trapped flux which has already existed in the bulk magnet by the former field applications. And the trapped field was gradually increased due to the decreasing heat generation by successive pulsed field applications [31]. The magnetic field distribution along the pole axis as a function of the gap is shown in Figure (5b). Although the magnetic field in the open space decreases with increasing the gap, the maximum magnetic field has reached 3.2 T at the center of the gap of 2 mm between pole surfaces. The field strength which each magnetic pole generates was kept constant at about 2.9 - 3.3 T at each bulk magnet surface of all the gap distances, showing sluggish enhancement to 3.2 T even when the gap of each 3-T magnetic pole comes close to 2 mm. This inferred that the magnetic flux invasion to the countering bulk magnet was restricted by the strong pinning effect, and the bulk magnets kept the trapped field as it was in spite of the approaching counter field each other [42]. Figure (6a) shows the distribution map between the face-to-face magnetic poles with a gap of 20 mm. The vacuum chambers are not indicated in the sketch. The map exhibits the characteristic property of the bulk magnet showing the strongest field spots just at the center of each bulk magnet surface. The highest value obtained in the 20-mm gap was 3.2 T by PFM called IMRA method which was operated by applying the field up to 6 T. One can utilize the strong magnetic field of 1.8 T at the center point of the gap space. As shown in Figure (6b), the demonstration of iron balls which are attracted by the generating field by the face-to-face bulk magnets shows a pair of very peculiar conical shapes which reflect the magnetic flux distributions well [36, 43].


A Review Article GM coolers Magnetic poles (Bulk magnets) Pulse coil

299

2.9-3.3 T (at the bulk magnet surface)

3.2T (at the center of 2mm gap)

Magnetic Field Bz（T)

4 3 2 1 0 -30 -20 -10

0

10

20

Position　（ｍｍ）

gap 2mm 10mm 20mm 30mm 30 40mm 50mm

Temperature control Helium compressors inside

(a)

(b)

Figure 5. Face-to-face HTS bulk magnet system activated by PFM bearing the pulse coil on one side (a), and the magnetic field distribution Bz as a function of various gap distances between the poles (b).

N pole 20mm

27

S pole Ferromagnetic Iron balls (3 mm)

Bulk

Bulk

REBCO

REBCO

18

9

Center line

Magnetic field direction

-5 0 5

Vacuum chamber 3.07T

1.8T 2.2T

(a)

3.21T

(b)

Figure 6. Distribution of Bz between the face-to-face magnetic poles (a), and the demonstration showing the conical shapes of iron balls attracted (b).

Attempt for Uniform Magnetic Space for NMR Application In order to expand the application area of bulk magnets, the magnet for the compact NMR analyzing system is on the way of the development. The NMR magnets in general require the extremely uniform and stable magnetic fields to detect the NMR signals in wide spaces. When we construct the hollow type bulk magnets, it becomes possible to realize the novel MRI/NMR magnets which are smaller than those in conventional



300

systems by far [40, 44]. In the trend, the authors have attempted to obtain uniform field spaces between the pole pieces of face-to-face settled magnet system referred above. As shown in Figure (7a), although the magnetic field distribution on bulk magnets exhibits conical shape up to 1.4 T, we deformed the shape to be concave shape, exhibiting 0.87 T at the center when we attached the ferromagnetic iron plate on the pole surface. The dimensions of the plate were 100 x 100 mm2 and 2 mm in thickness. As shown in Figure (7b), although the distribution exhibited a concave shape in the space near the pole surface, it gradually changed to be conical with increasing distance from the surface. One can see the flat line at the position of 15 mm from the pole surface. The best uniformity we obtained was 2,516 ppm (parts per million) at 0.477 T at the center of 60mm gap. The uniformity is estimated on the 2 x 2 mm2 plane normal to z-axis between the pole surfaces. Since the target of the R&D is settled as 1,500 ppm at more than 0.3 T, the data might be capable of detecting NMR signals in near future [45, 46].

Concave 0.87T HTS bulk magnet pole

Magnetic flux density (T)

1.4T HTS bulk magnet pole

Right-hand side pole (along x-axis)

1.4

Conical

10mm 15mm 20mm

1.2 1.0

0.8 0.6 0.4 0.2 0.0

Iron sheet 60x100xt2, Fe-C (