Progresses and challenges in the development of high-field solenoidal magnets based on RE123 coated conductors

TLDR: The goal is to illustrate a close synergetic relationship between evolution of high-field magnets and advancement in superconductor technology.

Abstract: Recent progresses in the second generation REBa2Cu3O7 − x (RE123) coated conductor (CC) have paved a way for the development of superconducting solenoids capable of generating fields well above 23.5 T, i.e. the limit of NbTi−Nb3Sn-based magnets. However, the RE123 magnet still poses several fundamental and engineering challenges. In this work we review the state-of-the-art of conductor and magnet technologies. The goal is to illustrate a close synergetic relationship between evolution of high-field magnets and advancement in superconductor technology. The paper is organized in three parts: (1) the basics of RE123 CC fabrication technique, including latest developments to improve conductor performance and production throughput; (2) critical issues and innovative design concepts for the RE123-based magnet; and (3) an overview of noteworthy ongoing magnet projects.

Figures

Figure 4 77-K Ic vs. strain plots: blue squares correspond to a- and b-axes of the Gd123 film oriented parallel to the tape length, red circles correspond to a- and b-axes of the Gd123 film rotated by 45° from the longitudinal direction [56].

Figure 3 (a) Normalized 77-K critical current vs. intrinsic strain plots of an Y123 coated conductor for selected fields applied parallel and perpendicular to the c-axis. (b) Normalized 4.2-K critical current vs. intr sic strain plots in selected magnetic fields in the range 2 -14T parallel to the c-axis. Reprinted from [52].

Figure 11 (a) No-Insulation (NI) HTS single pancake test coil; (b) Test coil mounted in the LHe experimental setup. Reprinted with permission from Hahn et al. Applied Superconductivity, IEEE Transactions on 22 4302405 [144], Copyright 2012, IEEE.

Figure 12 Over-current test results of an NI HTS test coil at 4.2 K. Reprinted with permission from Hahn et al. Applied Superconductivity, IEEE Transactions on 22 4302405 [144], Copyright 2012, IEEE.

Figure 13 Magnetic fields and shielding currents in superconducting tape: (a) external magnetic field Be and selfmagnetic field Bs for superconducting tape, (b) shielding (magnetization) current against Be, and (c) shielding current against Bs. Reprinted from [148].

Figure 8 Different configuration of bridge joints. Left: PD joint; right: OD joint.

Full text

1
Progresses and challenges in the
development of high-field magnets based
on RE123 coated conductors
Carmine Senatore
1
, Matteo Alessandrini
2
, Andrea Lucarelli
2
, Riccardo Tediosi
2
, Davide Uglietti
3
and Yukikazu
Iwasa
4
1
Département de Physique de la Matière Condensée (DPMC) and Département de Physique Appliquée (GAP), University of
Geneva, Geneva CH-1211, Switzerland
2
Bruker Biospin AG, Fällanden, Switzerland
3
EPFL-CRPP, Fusion Technology, 5232 Villigen-PSI, Switzerland
4
Francis Bitter Magnet Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Abstract
Recent progresses in the second generation REBa
2
Cu
3
O
7-x
(RE123) coated conductor (CC) have paved a way for the
development of superconducting magnets capable of generating fields well above 23.5 T, i.e. the limit of NbTi-
Nb
3
Sn-based magnets. However, the RE123 magnet still poses several fundamental and engineering challenges. In
this work we review the state-of-the-art of conductor and magnet technologies. The goal is to illustrate a close
synergetic relationship between evolution of high-field magnets and advancement in superconductor technology. The
paper is organised in three parts: 1) the basics of RE123 CC fabrication technique, including latest developments to
improve conductor performance and production throughput; 2) critical issues and innovative design concepts for the
RE123-based magnet; and 3) an overview of noteworthy ongoing magnet projects.
Contents
1.
Introduction ................................................................................................................................ 2
2.
RE123 coated conductors for magnet applications .................................................................... 2
2.1.
Preparation techniques: template texturing ................................................................................................. 4
2.1.1.
RABiTS template ................................................................................................................................... 4
2.1.2.
IBAD template ........................................................................................................................................ 4
2.2.
Preparation techniques: RE123 layer deposition ......................................................................................... 5
2.2.1.
Metal Organic Deposition (MOD) .......................................................................................................... 6
2.2.2.
Metal Organic Chemical Vapor Deposition (MOCVD) .......................................................................... 6
2.2.3.
Pulsed Laser Deposition (PLD) .............................................................................................................. 6
2.2.4.
Reactive Co-Evaporation (RCE) ............................................................................................................. 7
2.3.
Strategies to reduce the anisotropy of of critical current density (J
c
) .......................................................... 8
2.4.
Strategies to reduce hysteretic losses......................................................................................................... 10
2.5.
Effects of mechanical loads on I
c
.............................................................................................................. 10

2
2.5.1.
Reversible effects of strain ................................................................................................................... 10
2.5.2.
Irreversible limit under axial loads: influence of the conductor design ................................................ 12
2.5.3.
Transversal stress and delamination ..................................................................................................... 15
3.
Engineering issues in 2G HTS coil technology ........................................................................ 17
3.1.
Coil Technology ........................................................................................................................................ 17
3.1.1.
Winding Techniques ............................................................................................................................. 17
3.1.2.
Critical aspects of winding ................................................................................................................... 17
3.1.3.
Coil Impregnation ................................................................................................................................. 18
3.1.4.
Coil Insulation ...................................................................................................................................... 19
3.1.5.
Splicing ................................................................................................................................................. 20
3.1.6.
Stress Management ............................................................................................................................... 22
3.2.
Quench Protection ..................................................................................................................................... 23
3.2.1.
No-Insulation (NI) Coils ....................................................................................................................... 25
3.3.
Field quality............................................................................................................................................... 26
4.
Overview on Ongoing High-Field DC Magnet projects based on RE123 ............................... 29
5.
Conclusions .............................................................................................................................. 31
1. Introduction
Discovery in 1986 of high-temperature superconductivity (HTS) changed the attitude toward superconductivity of
many people, inside science and engineering and even outside. Because for the very first time the remarkable
potentials of superconductivity, much touted for so long, appeared much closer to the real world, at least as real as
readily accessible liquid nitrogen compared with remote and esoteric world of liquid helium. Although the 77-K
barrier was first broken by YBa
2
Cu
3
O
7-x
(YBCO) in the bulk form [1], YBCO was quickly followed by
(BiPb)
2
Sr
2
Ca
n
−1
Cu
n
O
2n+4
, known BSCCO that comes in two varieties, Bi2212 (n=2) and Bi2223 (n=3) [2]. From the
get-go, BSCCO fabricated in wire became the most eligible HTS for two important applications, power cables and
magnets. As have been long established by NbTi and Nb
3
Sn magnets, the preferred conductor form is indisputably
multifilamentary wire. Of the two varieties, only Bi2212 can be fabricated into multifilamentary wire. This is the
vital asset of Bi2212 that has kept it relevant from the beginning. For >1-GHz (>23.5 T) NMR magnets, Bi2212,
particularly with recent progress [3], is poised to become a viable alternative to the 2
nd
generation REBa
2
Cu
3
O
7-x
(RE123, RE = rare earth) coated conductor [4]. Today, RE123 is moving towards a goal of ever-expanding
possibilities and opportunities in a three-legged race with its most important partner, the magnet. This paper covers
challenges to and recent progress in the RE123 and high-field superconducting magnets. The paper first discusses
RE123 issues, including preparation techniques, strategies to reduce critical-current anisotropy, and stress effects on
critical current. Most of the second half is devoted on magnet technology, specifically on winding techniques,
quench protection, and field quality. The paper ends with an overview of 6 high-field magnets that rely partly or
wholly on RE123 coils.
2. RE123 coated conductors for magnet applications
In the last ten years the RE123 coated conductors have attracted most of the attention in the applied
superconductivity research. The reasons for this excitement are their high current densities in high magnetic fields,

3
their outstanding mechanical properties (with Hastelloy or stainless steel as substrate) and, above all, a prospect of
economically viable conductor prices.
The development of practical superconductors based on RE123 has revolved around the complexity of achieving
high critical current densities in polycrystalline materials. The problem lies in the build up of charge inhomogeneities
and strain at the grain boundaries (GBs), regions of mismatch between crystallites with misoriented crystalline axes,
that act as weak links that drastically limit the current flow [5]. Therefore, the critical current density through the
RE123 GBs falls off exponentially with increasing misorientation angle θ of adjacent crystallites for θ above 5°-10°
[5, 6]: it has been necessary to develop expensive and sophisticated crystallographic texture fabrication processes to
eliminate all but low-angle grain boundaries in useful conductor form. A variety of approaches have been
implemented at industrial level over the years, but all rely on two common features: a biaxially textured template,
consisting of a long and flexible tape-shaped metallic substrate coated with a multifunctional oxide barrier; and an
epitaxial RE123 layer. In addition, for environmental protection and thermal stabilization, an Ag layer a few-µm
thick and a thicker Cu layer complete the conductor.
At present, the commercial development of the so-called RE123 coated conductors focuses on three linked issues:
performance; price; and production length. On one hand manufacturers aim to improve the conductor performance
without increasing the production costs, on the other they are actively pursuing new routes to simplify the wire
architecture and increase the produced length and yield, with the goal to reduce the conductor price. The coated
conductor cost is mostly due to a large investment in the production infrastructure. The industry is therefore focusing
on to process longer lengths of input material in each production run, increase the useful output from the equipment
and thus lower the break-even price per unit length. The key ingredients to improve the performance/cost ratio are
three:
• Simplify the wire architecture, the production yield being the product of the yield for each production step;
• Perform buffer and RE123 deposition on large areas, followed by a slitting operation to size the conductor
to a final width;
• Increase the production line speed.
Kilometer lengths of coated conductors have been successfully demonstrated and lengths of 150+ m with a minimum
critical current of 350 A/cm-width at 77 K and in self-field are currently available. However, the present price of
~$200 /kA
m (at B = 2.5 T, T = 30 K) [7] is too high for widespread marketplace applications: power device
manufacturers (transformers, motors, SMES, cables, fault-current limiters) demand a price less than $50/kA
m.
In a high-field NbTi/Nb
3
Sn magnet, the operating current is often limited by wire strength (hoop stress), not by
critical current. Indeed, for high-field magnets in the field range 19 T-24 T, RE123 and Nb
3
Sn costs are rather
competitive. The reason is mechanical, i.e., an RE123 coil can operate at much higher electromechanical stresses
than its LTS counterpart, allowing it to fully exploit available high critical current density. Moreover, in a field above
~18 T, Nb
3
Sn has to be operated at 2.2 K, while RE123 at 4.2 K. Also note that the tape length required for power
equipment is large compared to that for a high-field magnet: hundreds of km of tape are needed for a 1 km-long
power cable, whereas continuous lengths of less than 1 km and total lengths typically of ~10 km are sufficient for a
high-field HTS coil.
The remaining part of the section describes the primary approaches for fabrication of a textured template and
deposition of a superconducting layer and addresses the latest developments in the manufacturing process of RE123
coated conductors.

4
2.1. Preparation techniques: template texturing
In commercial coated conductors, the textured template is created by either deforming the metal substrate with the
Rolling Assisted Biaxially Textured Substrate technology (RABiTS) [8] or by texturing the buffer layers by Ion
Beam Assisted Deposition (IBAD) [9]. The main features of these two technologies are briefly described below.
2.1.1. RABiTS template
In the RABiTS approach, cube texturing is achieved in a NiW substrate through a process of rolling and
recrystallization annealing [8]. Compared with pure Ni, NiW alloys exhibit reduced magnetization (Curie
temperatures of 335 K for Ni5at.%W and 630 K for Ni) and higher mechanical strength (yield strengths of ~150
MPa at room temperature for Ni5at.%W and ~70 MPa for Ni).
Ni5at.%W substrates are industrially prepared in the form of a flexible strip, typically 75-µm thick and more than 1-
km long. The X-ray full-width half-maximum (FWHM) of the diffraction peak referring to the in-plane texture of the
substrate along the entire production length is ∆φ ~ 6-7° [10]. The out-of-plane (∆χ) FWHM values average
approximately 7° [10] . After plasma cleaning to reduce surface roughness, epitaxial buffer layers must be deposited
to provide a lattice-matched surface ready for the growth of an HTS layer.
The first layer is typically Y
2
O
3
(AMSC, [10]) or CeO
2
(Sumitomo, [11]). This seed layer achieves an out-of-plane
texture that is sharper than the NiW substrate. This is a key step in the production of the RABiTS template. The
second yttrium-stabilized zirconia (YSZ) layer behaves as a blocking layer for the inter-diffusion of substrate atoms.
Top CeO
2
is then prepared to compensate the lattice mismatch between the superconductor and the YSZ layer. The
enhanced texture of the seed layer with respect to the substrate propagates throughout the subsequent YSZ and CeO
2
buffer layers. This template results in an improvement of ~3° in ∆χ and ~1° in ∆φ [12] compared with the NiW tape.
All buffer layers are grown by RF-sputtering. The layer thickness varies among manufacturers, but it is typically of
the order of 100 nm.
Further developments of the RABiTS technology are oriented towards substrates characterized by a reduced
ferromagnetic behaviour, in order to lower the hysteretic losses especially for power applications. To this end, higher
W-content NiW substrates with the qualified cube texture are being developed [13], but also new preparation
approaches for composite substrates are reported [14]. The concept of a composite substrate is to have a thin
biaxially textured Ni5at.%W layer on a non-magnetic Ni9at.%W base substrate. Although the biaxially textured
layer in this clad-type substrate is ferromagnetic, the overall magnetization is reduced.
2.1.2. IBAD template
The IBAD technique generates a preferred texture in the buffer layer, hence it does not require a textured substrate.
Polycrystalline Hastelloy or stainless steel tapes (typical 50-100 µm thick) are submitted to an electropolishing
process. This process reduces surface roughness from ~50 nm to less than 2 nm [15]. On the polished substrate
surface an Al
2
O
3
diffusion-barrier and an Y
2
O
3
seed layers are deposited by sputtering technique. The Al
2
O
3
and
Y
2
O
3
layers are 100-200 nm and 10-20 nm thick, respectively [16]. A biaxially textured MgO layer is then grown on
the Y
2
O
3
layer with the Ion Beam Assisted Deposition. This technique deposits MgO in a standard PLD geometry,
using a RF plasma source that assists ion beam. The assisting ion is typically Ar
+
with beam energy of ~1 keV and an
irradiating angle between 45° and 55° from the substrate normal.
Two mechanisms have been proposed to explain how ion beams affect the crystallographic orientation of growing
films. The variation of growth rate with crystal orientation is associated with the existence of crystallographic

5
directions along which ions easily penetrate into the crystal. Consequently, grains having these channeling directions
aligned parallel with the ion beam have a higher growth rate than nonaligned grains. The second mechanism is
related to the recrystallization on the substrate. The aligned grains are less damaged than nonaligned grain by the
ions. During the thermal spike associated with ion bombardment, the aligned grains can grow into their more
damaged surroundings, thereby reorienting the more damaged material [17]. Typically an appropriate texture is
achieved for a very thin layer (<10nm), though a homoepitaxial layer of MgO (~50nm) [15] is often added. The
typical in-plane texture ∆φ in the MgO layer is < 10° [16].
The next process may vary among manufacturers. At Fujikura, a 500-nm thick CeO
2
layer is deposited on MgO and
a sharp in-plane texture ∆φ=4° is obtained [18]. SuperPower, SuNAM and SuperOx grow a LaMnO
3
layer (20-30
nm) by sputtering atop the homoepitaxial MgO [15, 19, 20].
A variant of the IBAD process, the Alternating Beam Assisted Deposition (ABAD), has been developed by Bruker
[21]. In this approach, the substrate, first exposed to a molecular flow provided by a sputter source, acquires a few
nanometer-thick layer of YSZ. In the following step this layer undergoes an ion etching with an Ar
+
ion beam. After
several tens of deposition-etching cycles, a sufficiently high in-plane texture (with FWHM = 8-9°) is achieved in the
YSZ layer. A high rate pulsed laser deposition (HR-PLD) is then used for growing a ~0.1-µm thick CeO
2
cap layer.
The American manufacturer STI has focused on the development of a simplified IBAD architecture where the
unpolished Hastelloy substrates are planarized by a solution deposition process (SDP) [22], eliminating the need for
an electropolishing step. The idea is to coat the unpolished substrate with a liquid solution: the liquid surface tension
planarizes the free surface, resulting in thicker layers over valleys and thinner layers over peaks, overall, a smoother
surface than that with the underlying substrate. Specifically, several coatings of Y
2
O
3
and Y
2
O
3
-Al
2
O
3
mixtures are
used to planarize the substrate, leading to a surface roughness of less than ~2 nm. SDP has thus the advantage of
replacing three steps: electropolishing; barrier; and nucleation layer deposition. The material deposited by SDP is
then used as a bed layer for the IBAD and homoepitaxial MgO. In the STI template, HTS is deposited on the MgO
layer without the need for additional epitaxial buffers, such as LaMnO
3
or CeO
2
.
It should be stressed that for power applications both NiW and Hastelloy/Steel substrates can be used, while for high-
field magnets the substrate mechanical strength of Hastelloy trumps that of NiW. Indeed, this superior mechanical
strength makes the Hastelloy-substrate RE123 coated conductor a preferred option to all other HTS for high-field
(>20 T) magnets.
2.2. Preparation techniques: RE123 layer deposition
In commercial coated conductors multiple techniques for depositing epitaxial RE123 are presently being adopted,
with a goal of achieving a low-cost flexible process that is also easily scalable to mass industrial production. Table 1
lists key parameters of the coated conductors among manufacturers. The RE123 layer can be deposited either by
chemical routes, such as metal organic deposition (MOD) [23] and metal organic chemical vapor deposition
(MOCVD) [24], or by physical routes, such as pulsed laser deposition (PLD) [11, 16, 25, 26] and reactive co-
evaporation (RCE) [15, 22]. MOD and RCE are ex-situ processes incorporating two steps: deposition of the
precursors and subsequent conversion of the precursors into RE123. On the other hand, deposition and formation
occur simultaneously during in-situ processes such as PLD and MOCVD.
The basic principles defining these alternative approaches currently used at industrial level are summarized in the
following subsections.

Frequently asked questions

Q1. What have the authors contributed in "Progresses and challenges in the development of high-field magnets based on re123 coated conductors" ?

In this work the authors review the state-of-the-art of conductor and magnet technologies. The paper is organised in three parts: 1 ) the basics of RE123 CC fabrication technique, including latest developments to improve conductor performance and production throughput ; 2 ) critical issues and innovative design concepts for the RE123-based magnet ; and 3 ) an overview of noteworthy ongoing magnet projects. 

Q2. Why is the mechanical strength of the RE123 coated conductor compared with BSCCO tape?

Due to the large fractional section area of metallic substrates, mechanical strength is a big asset of RE123 coated conductor compared with BSCCO tape and wire. 

Q3. Why is RE123 the only high-temperature superconductor?

because of its current-carrying capacity and, equally importantly, high mechanical strength, RE123 is the only high-temperature superconductor applicable to high-field magnets. 

Q4. What is the effect of high stress on the composite tape?

High stresses have higher probabilities to quickly damage the ceramic compound, whereas large fatigue stress ratio can slowly generate cracks in the metals of the composite tape. 

Q5. What is the effect of adding metallic layers to ceramics?

It is known that the addition of metallic layers to ceramics greatly improves the fracture toughness of these brittle phases, because metallic layer acts as crack inhibiter/arrester. 

Q6. How many kilometers of coated conductors have been successfully demonstrated?

Kilometer lengths of coated conductors have been successfully demonstrated and lengths of 150+ m with a minimum critical current of 350 A/cm-width at 77 K and in self-field are currently available. 

Q7. What is the way to provide the insulation between pancakes?

In double-pancake coils the tape may be insulated only on one side (layer-to-layer insulation by mean of co-winding an insulating tape), with solid spacers providing the insulation between pancakes. 

Q8. What is the way to limit the strain on the superconducting film?

The conductor substrate metal should be strong enough (a high elastic modulus and a large elastic strain limit) to bear most of the hoop stress to limit the strain on the superconducting film. 

Q9. What are the main issues in the development of the so-called RE123 coated conductors?

At present, the commercial development of the so-called RE123 coated conductors focuses on three linked issues: performance; price; and production length. 

Q10. What is the disadvantage of keeping the RE123 film on the radially-outer side?

If the RE123 film is kept on the radially-outer side (tension), it can be a problem for magnets requiring small bending radii (small radius coil formers). 

Q11. What is the way to achieve a defect-free wire length?

Achieving long, defect-free wire lengths is a true challenge that forces manufacturers to simultaneously increase the batch length and improve the process yield (defect-free length / km of processed conductor). 

Q12. What is the impact of anisotropy on magnet design?

The impact of anisotropy on magnet design is very application-specific and depends from several factors like the geometry of the HTS coil itself (winding thickness, coil length, inner diameter), the presence, design and position of external magnetic sources (outserts) and the coil operating conditions (in particular temperature and ramping/cooling strategies). 

Q13. What is the length of conductor required for a DP coil?

a DP coil requires a continuous length of conductor that is considerably shorter than that required in a typical LW coil, e.g., even a 50 m long conductor may suffice for a DP coil. 

Q14. How can a laser beam be used to re-oxygenate Y123?

Haigun Lee has recently shown that the re-oxygenation of Y123 during annealing can be facilitated by opening micro-holes on the surface of the tapes with a laser beam. 

Q15. What is the way to speed up the NZPV in coated conductors?

A very interesting approach to speed up the NZPV in coated conductors is based on increased contact resistance between the superconducting layer and the stabiliser/substrate [119]. 

Q16. What is the method for restoring the oxygen content of the coil?

This approach consists in sintering two counter-facing ceramic layers in vacuum (>800°C) and restoring the optimal oxygen content by a special heat-treatment performed at a lower temperature. 

Q17. What is the effect of the a and b axes of the RE123 film?

On the other hand, critical current becomes almost insensitive to strain when coated conductors areprepared with a and b axes of the RE123 film rotated by 45° from the longitudinal direction (Fig. 4) [55, 56].